CONVERSATIONS IN SOIL TAXONOMY (ORIGINAL TR,.&NSCR|F'T~ONS OF TAPED CONVERSATIONS)

by

Guy D Smith

Compiled by an editorial committee a,

'I~,e Agronomy Departmen: of

Cornell University for the

So~i Management Support St.:vice

USD/ - Sf~

Ithaca, New York

.~996

7 ̧

~" "-2

2 . . =

. C

. % -

z -

Addendum to:

THE GUY SM|TH INTERVIEWS:

RATIONALE FOR CONCEPTS

tr~ SOIL TAXONOMY

by Guy O. Smith

Edited by

T.P,~. Forbes

Reviewed by

N. Ahmad J. Comerma H. Eswaran

K. Flach T.R. Forbes

B. Hajek W. Johnson

J. McClelland F.T. MiJqer J. Nichols J. Rourl,:e

R. Rust A. Van Wambeke

J. W ~ y

S'~d Management Support Services Soil Conservation Service

LL $. Dep~,rtment of Agriculture

New York State Co!le,ge of Agriculture and Life Sciences Cornei l University

Department of Agronomy

1986

SM:~S Technical Monograph No. i I

o -

. 1

"f Ib!e of Contents

Preface

Interview by Mike L. Leanly

Interview by J. Witty & R. Guthrie

Interview at the Agronomy Department at Corneil University

Interview at the Agronomy bepartn'e.'.zt at University of Minneso,m

Interview by H. E.qwaran

Lecture Given at the University of the West Indies

Interview at the Agronomy Department at Texas A & M University

Interviews b3: Coplar~ar staff & J. Comerma, Venezuela

ii

1

37

48

149

312

322

328

441

#rrr

Preface

Many papers have been published explaining the rationale for properties and class limits used in Soil T<:txonomy, a system of .soil classificalion for making and interpreting soil surveys (U.S. Department of Agrical~.ure, 1975) before and since its publication. Since 6"oil Taxonomy does not provide these rationale, many ~cientists f¢.lt that it wou!d be usefm to document the reasons for matiy of the decisions expla~nit~g the selection ~f properties and class limits.

The one person who ,,.'as fully conversant with the system and who co-ordinated it':; design was the late Dr. Guy D. Smith. !n i976, Dr. M. Leamy and s',aff of the Soil Bureau of Ne~, Zealand conducted a ser;,es of interviews with Dr. Smith. These interviews were published in the l~tewsletter of the New Zeal',,.nd l~i! Science ~ c i e t y and iater reprimed in ,Soil Survey Horizons. The considerable interest shown in these intervlie,vs was :he impetus necessary for the Soi' Management Suppor,' ~rvice~ (SMSS), established in October 1979, to continue this e f fo r t

In 1980 and 1981:, SMSS a l~nged a series of interviews at the Ur:iversiW of Ghent, Belgium. Cornell UnNersi ty, University of Minne,..eta, Texas A&M U,aiversiW, and with the Soil Conservation.,. Service. (SCS). Dr. Smith also travelleg to Venezt)ei. and Trinidad and w~.s interviewed by colleagues at institutions in these countries.

The format of the inte)views were similar at each place. All interested persons were invited and were free to ask questinns on all aspects e" Soil Taxonomy. However, the cGc-rdinator of the interviews at each 01ace also developed a list of majo r subject matter areas for discussion. Both the questions and answers were taped and reproduced.

Although the intent wa£, ~o CGV~,~ ~ much of Soil Taxonomy as possible, Dr. 5mith's failing health forced th , termination of the interviews in late 1981. Dr. Smith, did not have an opportunity to review the transcripts and co::sequeni,~y the Iranscriy>ts are reproduced with only ,ome e.ditorial changes. RecMzis ar.a advised to bear this in mind when they use :hese trar,.~cnpts.

The success of the interviews is a!so due to the large number cf persons who came to discuss with Dr. Guy D. Smlih. It is not possible to list .-all the names but we would like to recognize the main co-ordinators, wko -,~.~r-""

Dr. M. Leamy (New Zealand); Dr. R. Tavernier (Belgium); Dr. R. Ru,o; (Minnesota); Dr. B. A~len (Texas); Dr. A. Van WambeRe and Dr. M. G. Cti.~e (Ce, rnell); Dr. L. Wilding (Texas); Dr~ J. Comerm~ (V'ene:~,~ela), and Dr. N. Ahmad

" t g ~ (Trinidad). Stafff of .h,. So~l Conservation Service, particularly Dr. R. Arnotd, R. Gu,.ar~e (formerly SCS) and J. Witty (Washington, D.C.): J. Nichols (Texas); S. Riegen (Alaska) and F. Gilbert (New Ycrk) also contribmed to the interviews.

"/." k

ii

Dr. H. Eswaran put an extraordinary amount of work in transcribing a large ~et of origine, i tapes. These were at a later stage compiled, edited and indexed by Dr. T. Forbes, who also coordinated tb.e final publishing.

As ind~c, ated previously, ~he :,nterviews are not necessarily complete. There are still many more questions that could be ask,ed. However, this monograph serves to provide some aspects of the thinking that was behind t~,e formulation of the document. From this point of view, we hope ,,'his will be a useful documen+: Z~ all users of S-'~! Taxommly.

+ o .

- 1 1 1 -

l_,eamy Inter'view

Interview by Mike L. Leamy

July 10 - 20, 1980

Ghent, Belgium

l.eamy Interview

ntroduction

These conversations were commenced in 1977 when Dr. Smith was consultant in taxonomy to the New Zealand Soil Bureau. The conversations commenced with a discussion on the diagnostic horizons and the reasons for many of the parameters that had been used, and by the time Dr. Smith left New Zealar,.d, tiaey had not b',-.'er~ concluded. The continuation of these conversations took place in Ghent, Be,~giam in July 1980, and concentrates initially on a large number of questions which have beam colk-:ted world wide about many aspects of Soil Taxonomy. These questions have been derived from people who are using Soil Taxonomy to make or interpret soil surveys or to teach Soil Taxo~omy to students.

Question 1

The ~irst question w~-,:cn r., of concern to mar, y people who are using ,Soil Taxonomy to make soil surveys, is aboui the large number of taxa that is possible tc, identify within ~:~ mapping unit particularly at large scale soil surveys. Traditionally pedoIogists have regarded m:,.pping units as a homogeneous coacept for mapping purposes and while they have known that there is variation within that mapping unit, it has not been able to be defined. Now with Soil Taxonomy it is possible to def,int" the variations, and in some very detailed surveys; in New Zealand some mapping units on fl,~od plains may contain three or even f'~ur different Orders. This is a source of some concern particularly for younger pedolcgists and I would li!-e to ask you to comment on the advantages or disadvantgges and on the pF,:,losophy behind the mapping imit-taxonomic ani~ as expressed through Soil Taxonomy.

Smith:

It seems likely that at least some of tl':ose who are bothered by this problem have not read chapter 19, page ~07 of Soil Taxonomy. We stress there, that there is a distinction b,~tween the taxonomic unit, which is conceptual, and the mapping unit which portrays or attempts at least to pcrtray the real bodies of so)l that we find in the field. The limits of the polypedon, which is a. taxonomic unit, are controlled by r'xtural factors of soil formation. The limits of the mapping unit which attemDts to portray polypedons or associations or complexes of polypedons, a're controlled by another se: of tactors, namely me distribution and the size of the polypedons. Natural bodie.: that match the definition of polypedons are controlled by the same factors as the c~',ncept of the polypedons hut the lim;~B of the mapping unit are controlled by another set of factors which include both the scale of the map that we are making and the purposes for which we are making tile map. If the map is being made for very intensive land use, such as irrigated agt~.mture, we normally ~'at~t use a larger scale and we must show the variations in !he soil that are going to affect-the use of a parGcular sg,.ot l'or irrigation. ~ he same area being ma~,ped for extensive use, as range land, is going to be made at a very much smaller scale and we would ignore differences that we are required to show on the map made for irrigation. The problem arises when we atte~.pt to u_~ the same name f~r a taxonomic unit and a cartographic unit. The conc, eT~ts of the polypedon require the maker of the map to s~udy the mapping u~it and the Mnds of soil that it includes ratiie~ carefully so that he knows something ab.~out the actual variability of soil properties Within the are~. that he includes withir~ a single map delineation.

2

- 2 -

Leamy Interview

H.~..'ing done .',his, he must decide it~ putt,_'gg a name on the mapping delineation, on the kinds of variabilities he has and how these affect the use of the sol! for the probable uses that he can foresee. Soil differences that chaage the classification of the mapping unit from one Order to another, perhaps from Incep~isol to Mo'Aisol because of a difference of a f-.w centimeters in the thickness of the epipedon which changes it from ochric to mollie, may not be relevant to the use of the sol!. If, both the ~oil~ with and without the mollie epipedon have exactly t~e s: "~e family modifiers in the fa,-niiy name, it is unlikely that this difference is going to be relevant to any particular use. Therefore, in selecting the name, the maker gf the map may ""~c~ whichever of these taxa are morr~ extensive in the field. The user of the map is not ~arti,: ,~arly concerw:d x~'ith the taxono~ny, he is concerned with the interpretations that he is fur:~Lhed by the maker of th~ map. The ix.'n~ortant thing for the map maker is that he does not mislead the user of the map. If there are di~?ferences within the kinds of soil that are included within the map delineation th~. are sigaificat~t to the prospective uses of the soil, the maker of the map in selecting h2s name must then coesider the alternatives for names to reflect the presence of soils that behave in a significantly different manner. If the percentages are very small, he may either choose to negiect these in na~fing the unit or to indicate the locations of the contrasting soils with spot symboi~. Or ~f the variabilit~ is such that it ~.ffects management of the entire mapping unit he wii! ~robabl~ choose to name the mapping unit as a complex or as an association so that the map tkser i s warned that there are going te be specific problems in the use of that particular m~pping delineation. ~.ia the U.S. there are certain conventions thr~t are a,,~reed upon for reconciling the differences between the conceptual soil taxa in the cartographic mappin~ units. These standards i~ave changed in the past and will probably change again in the future. Soil surveys in other coumries will find it necessary if there are many workers to agree upon some standards so th'~'t ".here may be th~n be uniformity i n the nzming of the mapping units in the various :surveys that a're conducted concurrently.

Question 2

There is ~nother ques~io~ which i~ related in some way to the one you have just answered and it is question numbe~ I i (oC Rust). It appears to us that since Soil Taxonocc, y was adopted by the soil survey, mappers and corveL~tors have emphasized soil taxonomic units at the expense of natural soil landscap~ u n ~ , even more than they had b-=,~ore. Often ;he soil in the field needing to be classified at the seri~el l.,,:vel fits the concept of a widely used seri~s A according to ~ts stratigraphy and geemor~ho~ogy but fails to meet '.,;pine profile pror;erty requirement of that series o~' a requir~mzW, accumulated at higher categories of Soil Taxot~omy. Hiding beh~;nd taxadjuncts and ~:,ariance •does not .'-:ad~fy the correiat¢',~. Then using the descriptions ~md profile d~ta the soil is classified ~_t the family level, the list of series having that classification is eonsulte~l and series B is seieete~ even though its stratigraphy~ geomorphology and setting does not match the soil in the field. This results in a soil map ~hat is technically accurate according to Soil Taxonorny but does .~'~ot fit t_he landsca~. Do you recognize this problem and if so how do you propose to solve it? Stated another way, is an underlying assumption of Soil Taxonomy that soil profile pZOl~,erti~ at'~ more ~m~ortant than soil landscape properties?

Smit__.

The opening part of thks question appears to be a criticism more of the correlators application of Soil Taxonomy rather than of Soil Taxonomy itself. This is not a place for a comment o n t ha t part of the stat.e~.'nent rather than the question, The closing sentence in the question is a question that d e . r y e s aa answer and it does relate to the application of Soil Taxonomy to the ~ i l survey. Soil landscape properties are probably not well defined and since the man who submitted the question is not here to explain it, I am requ,:.red to ~iake my own

~: interpretation o f the question. I am ac~'~ming that an example of the problem referred to in this ~Uestion might be a land~cav'~,e in w h i c h we have a soil on an interfiuv in the upland~ and on a

• • ~ r - N i ~ ~

• : • " . . . . . k

~ , t - ~ ~ " ' "~ " "~ '" ~ " ' ~ ~ ' Loamy Interview

terrace bordering the strearn which has the same properties that have been used as diagnostic in I Soil Taxonomy. The one on the upland is presumably a different soil landscape than the one on II the terrace nearby. In this situation I believe Soil Taxonomy assumes that the soil properties are I more important then the difference in the geomorphic history of that particular landscape. I There are many areas tha.t are covered by loess in the midwestern states where the same loess, ] blankets the terraces of difterent levels and the uplands. In this situation Soil Taxonomy surely | assumes that the soil properties are more important than the geomorphological history. If there II is a probable difference ia some particular behavior of the soil on the terrace and the st, A on I the upland, in that perhaps on the terrace a well is more apt to find water than in the upland, i then Soil Taxm:omy is ,~,ery clea r that this is an appropriate use of a soil phase.

Question 3

The next question concerns> the classification of dry polar soils and arises from considerable activity by New Zealand soils scientists in the Antarctic over the last 10 to 15 years. The question that they have, now that they are generating considerable data about these soils, is how should they be classified.

Smith: "

At the time Soil Taxoaxomy was written, there was virtually no modern description of a dry pol~..r soil in the literat~.,re. The definition of the aridic soil moisture regime is such that there is ne provision made for a polar soil that has an aridic soil moisture regime and accumulated significant amounts of salt. A soil that never gets a.s warm as f;ve deg,'ees at 50 centimeter~ depth caanct be an Aridisol because it does not have an ar'dic soil moisture regime a~ad it never reaches th~ 5 degree l~ait during the year so that it cannot be dried more than half of zero time. The gap left between the definition of aridic, ustic and xeric soi~ moisture

regimes was deliberate. We have n.,') information about these soils that enable ~as to develop that part of the taxonomy and had we attempted to close that gap sp,, that there would be a place for every soil, we feared that ,the peddle, gist might attempt to ci~ssify the soil by simply applyi~g the def~,nitiens in Soil Taxonotx).y. ~'~ must be remembered th.~t classification involves not only the application of the rules to see where the soil fits in Soil Taxono~:y but equall)" importam"O,, it requires that the classifier study that classification to. see whether that is approl>riate. Many of the limits in Soil Taxonomj'. were selected to group the soils of the U.S. into classes that had some real mea.-Hng. The purno~ of classification is to p:;-t roger.her the objects that be~,ong together. How does the-classifier decide what things, do or do not belong together? The classification problem is no~,too difficult: he has the rule that the things, that belong together have common properties a:,d com:.,,,~o~_ behavior characteristics. A soil that has accumulated an appreciable conductivity u,'ader irrigation, may be capable of su~)porting at least one or even two crops a year u~der .rain-fed agric:altur,~ and yet the rules of taxonomy say that it is an Aridisol. Th[~ is obviously abs-lrd _if ~ e considers whether sach a soil th:~z accumulated its salts under irrigation add c~n loose. '2,era rea~Ci.ly, if they ~re leached to reclaim the soil from its sal=~ness. We-would then h a v e a soil that changes back and forth from an Aridisof to an lnceptisc~ ~, according to the year tb._at the )eacbdng ~s carried out. The absurdity of this sort of classification should be eppareet to an~or, e who is more concerned with putting the things that belong together into a taxon, than following the rules that are set by the limits of Soil Taxonomy. The classification of the polar soils is going to be det..*,rmined by this general principle that the th~ngs that belong toge~$er h ~ , e similar morpholog:.'es a~,'t similar behavior. We left th~s question hangia3 se that those who hr,,,#e studied the spits can propose a reasonable cla,~sifieation.

. ,

=' 4 - -

Leamy Interview

Question 4

The next series of questions concerns buried soils and they are not really questions but are comments which !ndicate that people have difficulty in handling the buried soil concept in Soil Taxonomy. Some of the comments ask for more explanation, others like one from Australia which is question number 43 and those put together by professor Rust state, we feel that Soil Taxonomy falls short of accounting for buried soil profiles, in particular those of the Murray- Murrzmbidgee-Darling river valleys which form such a large portion of the arable areas in this vast but arid continent (referring to Australia). It is clear that people require some further explanation as to how to handle buried soils.

Smith:

On rereading what Soil Taxonomy has to say about the use of thapto subgroups, it seems clear that more could be said about o,,r intent for its use. We have only a few thapto subgroups that we accepted in the U.S. and thest, all involve buried Histosols that came within the control section of alluvial derived soil. The use of the term 'thapto' is discussed on page 88 under the hesding "Names of Multiple Subgroups Intergrading between the two given Great Groups". The concept of a thapto subgroup was that of a particular kind of an intergrade although the name is listed in the table of extragrading terms, the footnote says that the thapto subgroups are not strictly extragrades. With this concept of thapto subgroups as a special kind of intergrade between different kind'..' of soil, the thapto subgroup cannot be used ff the buried soil has the same class'~ficatior, as the soil as the surface. It would be absurd to have a Haplargid that intergraded to a Haplargid, so that a Thapto Haplargidic HapI~rgid would be an odd name. We therefore have generally kept the use of the properties of a buried soil either at the family or at the series Ievel provided that the presence of the buried soil was relevant to some purpose of the soil survey. If the buried soil has a strongly contrasting particle size then it would normally show up at the family level. Th:,s would be in line with the rules for showing particle size in the fzmily levei. If the ~artic~e size distribution of the buried soil is so similar to that of the surface soil, that is the modern soil above the buried soil, then it would be possible to show the presence of the buried soil a t t h e series level. This should ~ done if the buried soil has sc~me relevant effect on the intended purposes of the soil survey. However, our purpose for making soil survey is rarely to show the geomorphic history of the soil. It is not uncommoa in arid regions that we have an Aridisoi buried by another Aridisol and one would have to be making a special sort of su~'vey with special purposes to find this relevant to show at a v e , ry high categoric level.

Similarly, the soils formed on volcanic ash and pumice normally have buried soils; very frequently one or more within the control section that we use for the Andepts or Andisols. Here if the buried soil is another soil formed in ash, the family level permits us to show the contrasting particle size distribution and the series level would permit showing a buried A but as we find this to be almost normal in soils from volcanic ash, we have generally kept such differences at the series level if they were relevant or we have disregarded them completely.

Question 5

~'2, ̧ _

E The next group of quest;o;~ reIaL', to the Mollisol Order and I have had inquiries from_,

New Zealand from Centra! and South America and there is a question in those co!lected by ,~h:~;~i expresses most of the concern that is expressed about this. 1 professor Rust (number 24) ;~' "~'-

would fike to know why se. much er~phasgg is placed on the presence of the moilic ep:,pedon at the highest level of classification. Thi~ seems to group very dissimilar soils at the Order level,

%',

.

Leamy Interview

when the only common features of these soils is a relatively thick dark surface with a high base status.

Guy Smith:

First, I should like to point out that the mollic epipedon is req)Hred for all Mollisols but is also permitted in a number of other Orders, ~ncluding Inceptisols, Alfisols, Ultksols, and Vertisols. The mollie epipedon is not the only common feature of the Mollisols. The Mollisols must have, not only a mollie epipedon but they must have a base san.,ration of more than 50 percent in all sub horizons below the epipedon and within the control section. The molfic epipedon is only required fcr MolIisoLS, permissible i~'~ four other Orders, but prohibited in Entisols at~d AridisoLs. The concept of the mollie epipedon is not only that of the dark colored surface horizon of the Chernozem. Rather it is the concept of a dark colored epipedon in which there has been decomposition of plant residues underground in the presence of considerable amounts of calcium. In developing the concepts of the Orders of Soil Taxonomy, we looked for some common feature that would group the soils of the former great soil groups of Prairie Soils or Brunizems, Chernozems, Chestnut soils, and Reddish Chestnut soils. These were soils that had formed under the influence of a dominantly grass vegetation. The only common features that we could find amongst these .,;oils were, the presence of a dark colored surface korizon of variable thickness and high base saturation. In the U.S., very commonly there was a horizon of acc~tmulation of calcium carbonates but this was not a universr, l feature because it was missing amongst the prairie soils or Brunizems. We had in the previous classification, that had been in use in the U.S., a suborder titled, "Dark Colored Soils of the Semiarid, 3ubhumid and Humid Grasslands." This suborder was a modification of the classification of Marbut in which he divided all soils imp the Pednlfers and Pedocals. In 1938 it was desired to group the prairie soils with the Chernozems on the basis of the .dark colored surface horizons and the grass vegetation. With this rather long traditional emphasis on the grouping of the grassland shits, it iz not surprising that when we develo0ed Soil Taxonomy we continued to give it an important place in the classification.

Nevertheless, we recognize that there were other dark colored soils that have low base satu~,tion and that it w = always possible in fact probable that many of these had received .':n application of lime adequ.*.te to ct:ange their former umbric ep.ipedon into a mollie epipedon. This is why the moUic epipedon is permitted in soils that normally have rather acid subsoil horizons. Having reached ~he decision to use the presence of the mollie epipedon and high base saturation ~t the definition of the Order, we still had some other so.;ls that did not have a ,grass vegetation but did have a mollie epipedon; Amongst these were the P.en~zinas, and some brown forest soils. These had been considered amongst the o.',b.ers, as intrazor~,al soils ancl there was no readily available order to put them in on the basis of their genesis alone. So we simply included them with the Mollis')ls as a seoarate suborder.

There ~re of course ~till remaining serious problems about th~ definition of the mollic epipedon. We have many ~oils that have formed under a swamp vegetation th',~t have a mollie epipedon and that are currently groul~ed with the poorly drained so~Is that had f,~)rmerly a grass and sedge vegetation. Therefore, the saborder of Aquolls has a much wider geogr~,phic distribution then do the Llstolls or the Udoll:L

Th.e second problem relating to the mollie epipedon is the limits for thickness° In a number of soils the normal thickness of the moiiic epipedon is just at the limit of 25 centimeters. This mzkes considerable tro~abie for a pedologist who is a purist and wants to classify everything on t;~e bmsis of whether or not it fits the definition of a mollie epipedon without regard to whether or not the d~fference of one or ';wo centimeter~ in thicknes:~ is relevant, to the purposes o f his soil surveys. We also have the problem of the soils witi~ mollie epipedons in the inter tropical region~. The, definitions of Soil Taxonomy are written primarily fo r the soils of the U.S. and otl~er tempe~-~te regions. We poin) out specifically that we have no good opportunity to test the elas~ffication of the soils :m the inter tropical reg;'~ns in the U.S. ..~nd. Zhis testing must be d o n e .~n other countries. We think: over time, th~.t some of these problems can be worked out through-~he help of the international committees on taxonomic problems.

- - 6 -

Leamy Interview

Question 6

The next question I would like to ask, concerns the use of subgroups and in particular many practicing soil surveyors in c0rrelators in countries outside the U.S. b,.~come a little uncertain as to whether it is !egitima~e to apply subgroups in a Great group where that subgroup is not defined for th~ Great group. Now I know there are some things called implied subgroups and we have had some direction on that, but I wouid like you to discuss the use of subgroup generally and whether you can apply a subgroup which has been defined for a particular great group in one Order to another great group in another Order if it is applicable.

Guy Smith-

I have mentioned elsewhere, the fact that classification involves more than the sole application of the rules of taxono~y to the classification of the particular soil. It also involves the consideration of whether or not the soils that ccme together under the definitions of Soil Taxonomy are soils that belotxg togeiI~er and that the soils that belong ~ogether are the soils that ha~e similar properties and similar beho.vior. The similar bel~avior may apply to one use but to m~other and if there are significant differences in any " ' ~'~" parucu,.., use, then at some categoric level, soils do not belong together. Now at the subgroup level we have in most definitions that the typic subgroup does or does not have ~roperties A, B, C, D~ etc. If a soil is like the tyF-ic except for A it belongs in an uquic subgroup and generally if it is like the typic except for B it belongs in another subgroup. We have provided a subgroup re: the soils like A with or without B or the soils like the typic except for B with or without A except; or course, these definitions are not mutually exclusive. But ~uch a soil might be found in another country and no subgroup is provided. It is like the ~ypie except for A and B and Soil Ta.,,,,nomy may provides for those that are t,q=e the typic except for one or the other, not for both. At this point then, the classifier is faced with a problem that he must create or propose a new s:abgroup or he must propose a tt'odification of the definition of the subgroups already proposed in Soil Ta.~.onomy. In making h~s decision on what to do about ~his subgroup that is not provided for, he must go back to the ~3~neral principie that one classifies the soils where they belong and this classification ;s based on the soii behavior rather than on the 9roperties. There may be to some, an alaparent contradiction in what I have jur;t stated. Some soils in different orders do not behave differently in any ~i~nificant manner. The soils form a continuum and the soils on one side e t a limit and the soils on the (~ther side of the limit, just barely d~ not have significantly different behavior. So tha'c we have, for soils in one great group and in another great group nc important difference hi interpretation. If both are close to the limits that are given in Soil Taxonomy. here is a problem where ~he clasrdfier must use some judgment and ~hould propose somz general sort of rule for ~ipplie:ation in other countries. Soil 7axonomy will not be useful ieternationally if there are no internation,.dly agreed definitions of the taxa.

Question 7

1

J i "" .

o= .~ .

I~ many active nation~.,i soil surveys outside the United States, it is imperative that decisions about mza be made with duedispatch so that the interpretations of the soil survey can be proceeded with and the benefit to. ~he users be realized. 1~ seems to me that in the c~.e of definition of =ubgroups which h~ve no~ been specified for in Soil Taxonomy; could well be delegated to correlation systems in ,.v~tions that had active soil .';urveys and that internation~i acceptance w,~)uld be ache.eyed by notification at regular interval,~. Such notit~cation being not fo r approval but for informat.~on. I do not see that there would be any great danger in thi~ proeedare as tong as national ¢orrelators followed the rules w~...:h you have outlined in answering this question. I wonder if you would like to comment on t~ds statement.

. j ,

' 7

Leamy Interview

Guy Smith:

My first comment would be that this problem is not restricted to subgroups but also occurs in great groups and conceivably could occur at higher categories than the great groups. For example, in New Zealand we found in a Dystrochrept with a placic horizon, a combination of horizons that is not provided for in Soil Taxonomy. It was necessary, then, for me to consider whether or not placic horizon had an effec~ on the interpretations and required recognition of some sot • of a Dys',rochrept that had a placic horizop, Well, anyone who has seen a placic horizon realizes that it interferes seriously with movement of water and penetration of roots and has an iml=~ortant effect on our interpretations so that this combination of horizons required recognition of a new taxon at some categoric level. The normal rule for Soil Taxonomy was that )~e presence of a pan, like a placic horizon or fragipan, w ~ recognized at the great group level. The combination of the ochric epipedon, the cambic horizon, and the placie horizon, not being provided for in Soil Taxonomy and requiring a new group if we were consistent wi:h the recognition of pans in other taxa throughout the taxonomy, implied that a new great group was required. H~wever, the ochric epipedon was always marginal to an umbric epipedon and we also had the combination of an umbric epipedQ, n, a cambic horizon and a placic horizon, and the differences in the thickness of the horizon, dark-colored enough for umbric, was always close to the necessary limit of 25 cm. It might be 20 em or it might be 30, but this was the normal range o f thic'_.~ness. Therefore;, I made a propo.~al in New Zealand that we not worry about the presence of an umbric or an ochric epipedon in the definition of ~his combination of Inceptist~ls with placic horizons. This then requirt, d z change in the definitions of Oehrept.~, and Umbrepts s~ that one great group or another had either an ochric or an umbric epipedon and a plaeic horizc,n. This kept together this group of soils that belonged together. Similar occurrences of unanticipated combinations of horiz~ms and p:operties are surely going to be fomad in all categories. The problem of the undefined subgroups is no different f~om that of the undef'~ned great groups except that having more subgroups than great groups, it may be more common.

Coming back to !:he problem of decisions about the need for new subgroups or great groups or suborders, I have stated elsewhere that the de~,ision is based on considerations of what soils belong together and that decisi~.~n ~s based on both the ~oil similarities in soil properties and similarities in soil behavior. TO amplify on the business of soil behavior I should like to comment that the ingerpretations are reflections of soil behavior and a significant difference in an interpretation for behavior unde.r one management system or another or one use or another is the basis for a decision that tl~a behavior is the same. If there is a difference in any significant interpretation under any management ~ystem or any use then we must conclude that the soils ao not belong together at some categoric l~evel. The distinction may belong at the series level or the family level or a much higher categoric level but the soils that belong together at ": great grc'ul~ level surely cannot all belong together at a ~,eries level. These distinctions go by steps and the decision does requi re so,:ne judgment and it does require same sort of international agreement if we are going ,.'G have an internationally useful Soil Taxonomy.

Question 8

~'~I~e next question is number 44 and those put together by Dr. Rust and is as follows: In the suborder or great group level keys, the relevance of one criterion in the exclusion process is not-uniform. The aquic ~uborder goes first in many orde,"s but no..), in Mollisols, the pals great

:: group;,~ first in Xerults, s ~ o n d Ln Ust~lts, and fourth in Aquults. Why is this?

- 8 -

Loamy Interview

Guy Smith:

There is an importartt distinct.~on between a taxonomy and a key. Both are classifications of a sort but a key is a!_mos'~ purely ~n artificial classification rather than a natural or taxonomic classification. The order in which the taxa appear it~ a key, in Soil Taxonomy at least, is based entirely on ease of eomp,'ehension e f the definitions. Consider first tiae aquic suborders. In most Orders, the soils with aquic sol7 water regimes and with the necessary qualifications for the definition of the aquic suborder~, n~mely low chromas and so on, is a common requi,'ement of the aquie suborde~s. In the Mol!i.sois the suborder of Albo!ls comes ahead of Aquo[!s and some of the Albolls have all of the ~'eouirements nece:~sary for an Aquoll hut some are not quite so wet but the presence of the ~Ibie horizon plus the indications of impeded drainage were conz:~ered more important ":~ the Albe, li~ than just the presence of the characteristics of poorly drained soils. To keep together in one taxon the :,3ils that belong together from their behavioral characteristics we wanted to porto!it the aoil drainage to range from, perhaps, imperfec;ly or somewhat poorly drained to poorly dra"ned. This was accompanied by the requirement for the presence of an albic horizon e.nd of an abrupt textural change between the albic horizon and the argillic horizon and it was much simpler to put these soils t'irst ~n the key because the key became much shorter than it would have been had we put the Aquolts ahead of the Albolls in the key. This is purely artificial and w ~ done to permit the shortest possible statements in the key. The same principle applied to the pale great groups in the Ultisols. If plinthite or a fragipan was present we wante~ to emphasize this in the taxonomy and in the constructiot., of the key it was much simpie, to put '~hese ahead of the pale great groups which did not have p~.inthite or fragipans.

I should point out that in Chapter 7, on page 91, the use of keys throughou~ 'he texl is discussed. We point out that the reader or the user should use ,'.he key first to the Or,nor and to then select the most prob~ble order that he can find for the classification of a particular kind of soil. He then goes to the ipage ;,ndicated in the key and at that point he will find a complete definit ion of the Order in terms of r.he properties of the order an~ the distinctions between tha~ Order and other Orders. If the soil that he is concerned with meets the requirements of that particular order definition, :Len he continues on to the key to the suborders. Again he selects the most probable suborder', turret', to the indicated page, and then checks the particular soil against the definition of that suborder and so on down the line through the keys to the complete definitions of the various taxa.

Quest:ion 9

. , . .

Question ~50 compiled by Feof~sor Rus*= Is an organized order-by-order, subor.:~er-by- suborder, etc., table of differen'~i~e within each category of the taxozomy poss'ble that would portray the logic of the system ~,:s it waz developed? And the comment is made: Similar tables are given using the nomencl~Jre but none seems to have been devised to portray the logic.

Guy Smith

It might be possible to develop a table to portray the logic of Soil Taxa'tomy although this has not been attempted. W~ have discussed the ,logic of Soil Taxonomy in several pla.,:'.es. Marlin Clin.,r. has discussed it in zevere.i ~apers (Cline., 1949, 1961, 1975). i have at least two papers in whic~ the logic is discussed, in ~my lectures on soil el~s;,fication (Smith, 1965) and in another 9aper reference to ~ added later. Cline (1949) h~ls pointed out that if the full reasons f o r this selection of' a particular differentia.e are given, then the users are inel.ined to pay attention to ~/ose reasorw~ which i~vo!ve assumptions about soil ge~esis and so to the genetic asrumpzior~s than to the defin?dow;. If he does so, it blinds him to the possibi.lity that there are error~ in these ~ssumptions .~_ad i t tends to freeze., the "taxonomy i=~ a form that is no~ as good as

Loamy Interview

~:t should be. Cain, (referen~ce to be added later) has pointed out that, i:~ the botanical and in the zoological taxonomies tb~ere i5 t'ae phylogenetic taxonomies in which there is an assumption that a particular character that distinguishes ~ family arises only once in the course of evolution and the fewer the sim';IaritJ',es they find between plants or animals, the older this characteristic arose. This is just an assumption he goes on to ~ay and may not be correct. It may h8ve arisen independently at different times but the phylogenetie classification makes this other assumption and therefore it blinds the taxonomi.~ts in biology to this assumption which may or may not be correct and tends to freeze the taxot~omy in an imperfect form. Therefore, in the development of Soil Taxonomy we cure, fully hid trtost of our assumptions about the genesis of the various diagnostic properties that we have used in classifying the soi l This was hidden to prevent the freezing of the taxonomy into a s.t.erile system based on some genetic assumptions that might or might not be correct. Whether or not the table could be developed that reflected ~.~1 of these assumptions cannot be ~;een until someone tries to do it. However, the full intent is that this shall not be done so that ',he t~uture ~axonomists will not pay more attention to the faulty assumptions that we m~,ke today than to the definitions. The definitions if they do not work can be corrected. ":'he assumptions .if they do not work are more difficult to correct.

Question 10

Question #47 and those compiled by Professo: Rust: the concept of lnceptisols seems to be one of the more difficult to accels:. Can you give some background on the Order?

Guy Smith:

During the development of Soil Taxonomy certain groups of so.;t,3 appeared with many .:haracteristics or a few chr,-racteristics that were common that seemed to be closely related enough to just ify recognition as an order. For example the Vertisols, with their expanding clays ~,ad their occasional or frequent dry s,.~asons, with their w~de cracks, constituted a group of soils that it seems should be rec~,,gnized as distinct from other kinds of soils. Similarly, the Mollisols with their mollie epipedons and high base status, seem to warrant the recognition of their own order, the Entiso!s laek~np~ ~ 7 diagnostic horizons seem worthy of recognition as an order and so on. Eventually we h~,d nine apparently satisfactory groups of soils that could be used to recognize orders. However, every taxonomy hr~ a waste basket. When we finished with the nine Orders we s'dll had many soils left over that appeared better not grouped with any of the other f3rders. We tried, for example~ to group soils with cambic horizons and with argillic horizons in various approximations and none of the groupings seemed to be satisfactory. From a geneti'." point of view, one could group some of the lnceptisols with the Alfisols on the basis that they are going to develop into Alfisols o~ar time but we had to reject this kind of an assumption on the bas:..~ that the lnceptisols, being weakly developed, might develop into Alfisols o r Ultisols over time but t)n the other hand if erosion exceeded the rate of soil development they might develop it~to Entisols. It was not possible to group the soils on the basis of a genetic assumption th~:t over time they would develop into another kind of soil. So ~ !neeptisols repr.:sented the. kind of soils that ;iid not seem to fit into any other order. Over time no,, we have concluded that the suborder of Andepts has enough properties in common that they should be recognized as an eleventh Order. In time, the:',~ may be other Orders cut out of the suborder of lncept/~ols bat this is a matter of fu:~are knowledge rather than the ~r.?.sent .t~owledge that we had at the time we developed Sod Taxonomy.

- 1 0 -

Leamy Interview

Question 11

The next question concerns soil temperature regimes and it is #45 compiled by Professor Rust how the criteria for soil temperature regimes, that is, 0-8, 8-15, 15-22, and greater than 22 degrees C soil temperature are based on temperature-crop needs. Is the concept of zonality based on commercial crops valid in a natural system?

Guy Smith:

Well, before work on the development of Soil Taxonomy was started, it was recognized that the concepts of zonality arzd intrazonality were not tenable :., a natural classification because they were not based on soil properties. That is, not based on the propert;.es of the soils that were being classified. I t was necessary to classify the soil as zonal or intrazonal on the basis of properties of other ~oils zhan those being classified. Having recognized that soils could not be classified as zonal or intr~.zonal on the basis of their own properties, one had to fi. ~d substitutes for the highest category. The use of soil moisture and soi!~ tempera'~ure was a natural substitute for the concept of zonzl and intrazonal soils. In general the soils of a given region with the same rainfall have roughly similar soil moisture and soil temperature regimes so thaL with the exception of the coils w~th aquic moisture regimes, one had a sort of su~.~.zitute for zonality that was based on the propertie~ of the soils being classified. The soil temperature and soil moisture regimes were useful for classifying soils f~-om the too down in .~ descending order.

To classify the soils on an ascending order, meant grour, ing soil series into nar, ural ~.axa c,n the basis of their affinities. However, there were too many thousands of serges to be comprel~.nded by any one individual so th~qt the system as it developed w,~s a sort of compromise that soils were subdivided from the top down on the basis of certain properties and the classes that resulted were tested by examining the nature of the series that were grouped by the criteria from the top down. So the development of the taxonomy was a compromise classifying by subdivision or clas~ifyiog by combinations.

To understand the prob!ems that" were involved in classifying the soils by groupings of series one must understand that there were thousands of series (:oo many to be understood, by ~.,ny single individual) and that tt~e series had been in use for many years and had been tested by use and found to be; useftd. The newspaper advertisements in Iowa, for example, would advertise a farm for sale p~- 160 acre~ of C~r ington loam. This w ~ a use of the series name. The bayer and the seller got information from the use of tee name Carringtor,.. t'-ne highway engineers used the series as a basis for planning their secondary road construction. When the highway engineers heard tibet we were c~eveloping a new system of taxonomy, they required me to appear before the Highway Research Board to explain what we were going to do with t~,e soil series because they wan:ed ,.hem retained to the maximum extent possible and they were placated when I ~xplained that i.~ was our goal to split as few series as po3sible and only, when the split seemed to warrant a~ improvement in our interl~retations. When the criteria proposed in the earlier approximations' were e.xamined by seein,q bow the series were grouped, I received repeated complaints that this ~s apt good because this splits our series; the' fioal was to retain the se~es as ne~.rly as possible with their previous use. However, the series were not defined on .~he basis of temperature er moL~t)~r.~. These- were mferre0 characteristics and related to the series but not appearing in the series de,fh~&dons. Where the type of farming changed we made different interpretations. For ~xam.01e, t~e interpretations for soils cropped to cot-'.on wzre not the san:~ interpretations that we made for ,:oils cropped to maize G~ to spring wheat. Therefore. the series normally changed with the ,,-,he..,,. of farming. How it happens that the limit between the Cotton Belt and .the Corn Belt, between the Cotton Belt and the Winter Wheat Belt, between t h e Red Desert soils and the Gray Desert soils was the sam~:,, always a t 15 degrees C mean annual temperature. Th,~refo~'e~ this w~s a natural limit that did not split series. The Red Desert-Gray Desert separation w~z b~ed on the natural vegetation--creosote bush--being p~..sen): in the Red Desert and absent in the Gray Desert. If one studies the general soil map of the United States that w a s puollsh~_fi! In :he 193g Year Book of Agriculture - Soils and Men - it is immediately obvious that the boundary between Red-Yellow Podzolic soils and Gray-Brown Podz~lie soils follows the ~5 degree C soil temperature isotherms. This was not based on the

. -

- ! 1

Le~my Interview

type of farming because we currently have in the Alfisols, the recent soils on loess along the Mississippi Valley, although these were previously called Red-Yellow Podzolic soils and now they are thermic Hapludalfs and so oa. While the correlation is imperfect, the differences in type of natural vegetation were rather apparent but with an imperfect correlation between the distinction between thermic and merle soil temperature regimes.

It is impossible to use the naturai vegetation as a basis for cla~;ifying soils >ecause many soils have as theiz natural vegetation, commercial cultivated crops. Examples might be tee soils of the irrigated valleys of the Nile, or Tigres, or the Euphrates where the sediments hr~ve accumulated and the origi~ml soit ~ buried deeply below the present control section. The only vegetatior, -:hat has grown on the.s¢ so;is has been commercial crops. Rice, cotton, for example, in Southeastern Asia, in the U.S. we have similar situations on floodplains where the sediments have accumulated under cultivation and the originz,! soil is now deeply buried, perhaps to depths of 2 or 3 or 4 meters and ~he only vegetations these soils have had may be corn or cotton. These are their natural vegetation.

There are similar changes in type- of farming and in vegetation that cross the country and the 8 degree isotherm and at the 22 degree isotherm. The limit between the Cora Belt and the smzi! grains or the corn grown for silage comes at 8 degrees. The limit between winter wheat and s~ring wheat comes at .8 degrec~. The limit in the northe~tern states in New England, where we change from 'sol brun ~cides' or Dystrochrepts to Spodosols, comes at 8 degrees. So the series changed agairt zt 8 degrees across the country until one reached the Aridisols. However, there are few aeries of Ar';disols in the frigid zone; so that the splitting of series there was not of serious consequeace.

The limit of 22 degrees h3 the eastern part of the United States separates the citrus belt and the wiater vegetable belt from the other soils and again we had other series. So the use of the particular limits of 22 , 15, and 8, produced the least possible disturbance of the soii series. It coincided witi.,, the gener~,.l but not u.~iversai changes in the natural vegetation, where the natural vegetation could be determi~.ed.

In the tropics where we h~ve isotemp-'~rature regi~t~es, the natural vegetatio, frequently is not possib!e to determine. "f'he ecologists are still arguing about the origin of the savannahs in the tropics. The isotemp¢~ratuwe limits were selected for convenience to have the same limits as the mhers, mainly 22, 15, 8, for convenience of the user of taxonomy. We felt he could remember one set of limits muci~, more easily than he could two. The limit of 8 degrees for i, sofrigid from isomesic was wrong and suggestions have been made to change it. The limit of cultivation in the inter tropi~.:al regions has a mean annual temperature of the soil of about 10 degrees rather than 8.

It seems important, in ~, soil survey that is made to facilitate interpretations as well as mapping, that there be some relation between potentials for cultivated crops and the soil properties. We attempted in drawing the limits betwee~ the Aridiso;z and other soils to draw the limit between what could ~ ~:ultivated without irrigation and what could not. In the case of the isofrigid ~emperatures we would again g~nt to draw the limit between what can be cultivated and what eannoi becau.ge of nightly frosts.

Questions 12

The next question is ~.~-~a6_ on L ~ e compiled_ by Professor Rust. The classification of ~ost soils, possibly cxcepting the Oxisols, is based on morphology to a depth of ~_bout 1 meter. On the other hand, many applications i¢~ engineering and agriculture require info~.-:r~ation on materials below this de~th. How do we reconcile this problem?

2/ - i 2 -

?

Loamy Interview

For the most part, the control ~ection for the classification of soils at the family category, stops at a meter. The control section for series is permitted to run below a meter. If there are significant differences below the depth of 1 meter and above the depth of 2 meters, for the most part the classif.i ation would be reflected at the level of the soil series. Significant differences at this depth must be shown by some means for interpretations. If the differences occur below 2 meters, the man making the soil survey will have relatively few observations compared to his observation in the surface meter, which he can examine readily with a soil auger. Differences below the dep~th of 2 meters also need to be reflected in the mapping units if they are significant to the anticipated uses of the soil. However', these differences would necessarily be used as phases ra~her than as series or family differentiae. It L~ important that any difterence at any depth be shown at some categoric level or as a phase, if they affect t~e anticipated uses. However, the difference at a depth of 6 or 8 meters requires a power drill to determine and one has relatively few observations and the p!aase is about the only possible way to show these differences.

Question 13

The next series of question~ concern particular queries that we have received from people working with Soil Taxonomy throughout the world, about different aspects of the diagnostic horizons. The first one concerns the oxic horizon and we have the san,~ query from two sources question number 19 and those compiled by Professor Rust and also a question from a soil survey people in Thailand. The Thai people ask why the oxic horizon must have n ore than 15 percent clay? They commentecl ,:hat some soils in Thailand have red color, diffuse boundary, very high porosity b,.~t clo not have weatherable minerals, or clay coatings and the sub horizons fits all the requirements for oxic horizon except the clay parameter and they consider ,'hat it is not comfortable to identify this horizon as a cambic horizon because the soil is old and highly weathered. The questifm compiled by Professor Rust reads as follows: we have found sandy soils with about 5 percent clay ,';o~tent that have all the properties of Oxisols except the clay content and really the~e pro, pea'ties give them very specia~ nature and behavior. I do not really know the ,'easc, ns for a estricdng oxic horizon to horizons with ?~reater than 15 pe~-cent clay; there are extensive z~reas with s, andy soils that have, besides the property of sandy soils, marked characteristics that identify t,~em a~ Oxisols. Should we not permit very sandy soils in the order of Oxisols?

Guy Smith:

It would be possibte to per~lit ~.he oxic horizon to have a texture of ss.r.~d and te. lack all weatherable minerals. In this ca~e, the horizon would consist Gf quartz, f~ee oxides, perhaps traces of 1:1 lattice clays. Hov-aver, the Oxisois grade into the Quartzipsamments and if we ;nclude in Oxisols, soils with 5 percent clay then we must find some limit be,.'w~.en 5 and 5/10 percent clay to stop the oxic horizon a_~.a go it, to Quartzipsamments because Quartzipsamments are frequently highly weathered and the clay fraction consists of kaolinite and free oxides. There must be some sot-t of limit between the oxic horizon and the Quar'.zipsamments because on the landscape they grade one into another particularly in the Zaire and c~ther par;s of SouL3.1rn Africa.

I .have proposed the., • d;.,~ limit on clay be dropped and that a iimit on texture be substitm,~l. Namely the oxic horizon should have a sandy loam texture and the Quartzipsamments should have a ,.:andy text,ire. The taxonomy provides for ogic subgroups of the Quartzipsa_~men'.s and one wi thin the landscape go from an Oxisol to an Oxic Quartzipsamment and finally to a Typic Quartzips?,mment in which the ra~d grains are largely un coated. The clay limit was ir,~serted o.r:Jginaily in order to make a bteak between the sandy

- 1 3 -

Leamy lntervLw

loams and the sands or loamy sands on the assumption that the highly weathered :';oEs have virtually no silt. However, we have found a number of soils that have no weatherable minerals that have less than 15 percent clay actually 10 to 12, , d have cmough silt that they have a sandy loam texture rather tha~ ~, sand or loamy sand texture. To classify these soils as Quartzipsamments are impossible because of the sand? loam texture. To classify them as Oxisols is impossible according to the ',limits in Soil Taxonomy. They become orphans, and knowing something about the soils it seems obvious they belong L~etter with ihe Oxisols ~hen they do with the Entisols which is where they would have to be p,..tt if :hey had no diagnostic horizon. We have found such soils in Venezuela, and it seems very -~ikely that they will also occur in other parts of the wo~:Id. For the most part the Quartzipsamments have more than 99 percent unweatherable mi-mr~ls in the sil~, and sand fraction although the limit in taxonomy is set at 95 percent. They t ,present soils that may be very recent in origin, occurring on coastal dunes where the sands o~ ~he beach are ahnost pure quartz. They o:cur on very old landscapes where the sands have been in place for a lo~g 'time and have had all of the weatherable minerals removed. They al.o occur as greatly over-thickened albic horizons with an underlying spodic horizon zhat is more titan two meters deep. Most of such soils with the thick albic horizon or those on the recent dunes, are almost totally lacking in clay and dominated by quartz. The intergrades between the Oxisois zhat have a sandy loam oxic horizon and the Quartzipsamments, that are almost completely lacking in clay, mast find some place in the t~xonom3. T~. Psamments were distinguished from other soils on the grounds that they have some very specific physical properties. They are, when dry, subject to blowing and drifting. When dry they are also very difficult to traverse with wheeled vehicles. The Oxido~s on the other hand do not have these specific properties. So we need a limit somewhere between the Oxiso',s and the Psamments including Quartzipsamments that is based on the point at which we begin to develop these particular propertie.*; of psamments. It seemed reasonable to us when we developed the taxonomy, since ~.here is a continual gradation between Oxisols ^nd Quartzips~ml~.ents, to bare some limit that recognize the point at which we begin to develop trafficability and blowing problems. This was the basic reason for the 15 percent clay limit, tvhich we know now was wrong because the presence oi the appreciable silt plus clay does not produce the peculiar properties of the Psamments.

We prepared, about 10 years ago, a manual of field soil survey investig~fio~ts of she,ring the things that the field men could d~ in their offices without requiring the ~.xistence of services from a laboratory. One of the tests that we described i. ~, one for the estimation of the pea'cent age of quartz in the ~'md fraction. Thi~ is b,.qsed on covering the sand fraction with a liquid that has the same refractive i,:~dex as quartz. When one does this the quar~:z becomes invisible and the other kinds of sands remain in plain sight in the liquid,

Question 14

Question number 16 ¢o~,niled by Professor Rus~, reads aJ follows: Argillic horizons are accepted in profiles which belong to ~,he order of Aridisols, MollLso!s~ and Vertisols. Other pro~,erties (moisture ~gim~, moll~c e~ipedon, s~:,odic horizon, etc) were con.~ideced more important and ~iven me,re diagnostic w~ight to create O;de-s than the argillic horizons. This was not the case when the oxic hori~,n was cons':dered, wh.v'~ The 7th approximation (Soil Survey Staff, 1960) gave ~he promi,,nence in the key for soil Oraers, ~'o either the argiPAc ~t' ~he oxic horizon. Finally, in S¢,il Taxor~.my (Soil Survey Staff, 19"/5) the ar~illic was given priority over the oxic. Why'?. Q ~ t i o n nurnl~er 12 compiled b~' Professor Rust aiso fe~ates ~o the same qaestion as to wh3' more weight i~ u'~e soil key ~ given to the argillic ~han the oxic and asks if there is a reasoz, for ~hat.

1 4 -

Leamy L, aerview

Guy Smith:

I should point out firs~ that the emghasis to the argillic horizon over the oxic horizon applies only to the soils in which there is an argillic horizon overlying zhe oxic horizon. An argillic horizon underlying an o~:ic hc,~ir'~= is not g r o ~ d s for keeping a soil out of Oxisols. We generally use the principle in devel,opmg Soil Taxonomy that, if we have tw.?, subsurface diagnostic horizons in the soil, the preference is given at the higher ~ category to the h~rizz~n nearest the .~-~rface. Thus 'the soil wid: both a spocJic horizon an,~ an a,'gillic horizon is normally classified as a Spodosol ~ecauz<, the spodic horizon overfies the argillic horizon and the assm'nption is, that the more recent" pc~,,ces.~.,, th.~t dom.;r, ate in the genesis in the soil produce the diagnostic horizons elo'.er to the surface than the older process, which produce ~:he diagnostic horizon at a ~reater d~pth. T~iz assumption is consistently used in the various Orders where we have the two ,3r more diagr.~-,~,~.ic subsurface horizons. There is no distinction between the use of the argillic horizor~ in s~-';dosols and in OxisoB. I suspec~ these questions abo~t the argillic horizon's significance to ::r;,'~ .el?.ssificat.~on arise from a f,.',ilure to read carefully the full text of the discussion of the argiil~c horizon. On page 20 under the heading, 'Significance to Soil Classification', there appears this statement "It i3 stressed that the argillic h o r i : ~ is no more important ~o soil c~ass.;.f[c~'tt~on :~nd to soil genesis ,.hart many other hori,.ons. It has been used at a higher categoric !ev,ei in some parts of the system only beca~,,~se tha~ use has produced groupings of soils that h~ ~ .* ,he ]argeet number of common proi>~rties that are impo~tant to use of lne soils."

Question 15

The next question com,e-~ from Thai;and and concerns a particular ~tu~tion in which you ~,aay find an apparent ~gi ; l ic horizon. .Soils developed on foot hill ,;lopes consists of rock fragments not cock s t~c tu re and clay coatinga can be observed especially around the rock fragments. Should this be idet~tif.~_l as an argillic horiron even thougt: the soil occurs on th~ relatively young landsc~.pe?

Guy Smith:

It is very common ia sgiis tha~. b2ve a skeletal or fragmental parti~e~¢, size i3istribt,.tion to find coatings of clay on ',he ~'ocks. Soil Taxonomy specifies the im,~'ortance of clay coatings, clay ~kin~ on edge and i'./~ r ~ r ~ and .in the t;keletai and the fr~.gmental p,,rticle ,~ize classes, one does not ordinarily find i ~ because the stt'ucture is controlled by the rock fr:igment;, and i:, fragmental particle size classes I:eea~:6e ~here ~re no Ix~ds and there a~'e no pores other t~an the large ones that are not filled w~h f~ne earth. The coatings of clay on f~e T,_~ks in skeletal and fragmental particles s;~z~ c¢~'ges ind.:.cate that the clay has b~en moved, as a rule, becao~e if one studies the coarse f~gments~ ~'ou do not find evidence, oi ,veathe.ring sufficie,:at t~, produce the c~ay by weath~ring in place. I have seen in Norway, ~:~d in Maine, skeletal ~r fragmental particle classes, with ~lay eoa~ngs ~ t exteneled to a d~:pth of ~'nore than 5 meters. This d~oes indicate that the clay~ are ia t ~ a s i t but i~. does not indicate an accumulation rf tr~.~siocated cla$,. In the majority of ~:e soils b.~, my ~xperience, in which one claims cr~ticgs of clay on the rocks, there is not a suff[cier~t ~ccumulation of clay ~<~ satisfy,, the requirements of ~ argiilic horizon. "~he clays seem in b~,. ";n transit ~n l~v in8 the soil c:~mple~.ely *~ut not really accqmulating. The re fo~ Soi~; Taxc,:~omy refers to, accm'~ular:on of clay in clays ~kins. in pores and on gmda but it does ~ , t refer ~.o accumulations on rr~cks. It ~-,,s the 'intent that the accumulation of clays oa ro~.ks would not be considered adequate for recognition of an argillic horn, ann.

- i 5 -

Leamy Interview

Question 16

The next question concerns the difficulty of recognizing ','n argillic horizon in which soils with a high clay content, say ~0 ~,~ercent clay throughout the solum; it is very difficult to identify clay skins in such ~,nater]al and the parameter recommended in Soil Taxonomy is the ratio of fine clay to coarse clay, which is not easily recognizable in the field. The question is, are there any field recog~,ition testa which might help the fie~dman to identify an argillic horizon in such soils?

Gt_jy Smith:

I know of no field clues ~ a t can be used generally to recognize argillic horizons in soils with very high clay conterr~:s. If !he eluvial horizon has 80 percent clay and the illuvial horizon, 90 percent clay, the dist~nc~ion ~a~ be recognized by a very oxpe~'ienced pedologist. However, the beginners will not recognize the d~tinction between horizons that have 10 percent differences of clay when ~he clay cerement ~s so high. The things that one may see in the field would be the colors of the p~,8 face~ [n such soils. Not much else c~n be observed. To m~.ke interpretations of the prese.rzce ~r ~bsenoe of .,rgillic horizon some laboratory studies would be reedtired. The ratio of fine/coarse cizy is erie of the most important. The difference in clay content in such soils is commonly due to a difference in gedimentation ratl'.er than pedogenesis. Ho,~ever in a particular su'zvey ~,rea, with some laboratory armlysi~ of the coarse to f i re clay ratio one can judge that tb, ere ~t~-z,u~d be an argillic ho:izon or should not. Some benchmark studies in such soils it~ the la[,~I)ratory would be essential in order to have confidence of ti,e pzesence or absence of r~n argil.~ic h~.)r~zon. These benchmark ~tudies mu~: be on a ~urvey area basis. No general s~.~ement is i:~s~.ble.

Question 17

The nex,', question is number 34 and is comp,~led by Professor Rust and ~ise~ the whole question of the use of soil t em~ra tu re regimes at categoric levels higher than the family. The question concerns the use of trorpic gre3t groups instead of tropic subgroups. For example, if an Aquult ~,ith an isohypotkorm~c ternpc.~ture regimes has plinthite, a fragipan, an abrupt textural change, G¢ a 'pale' curve, it falls into a particular great group regardless of the temperature regime. While in the same general landscape a so~! without any of the al~ove features would be classified as a Typic Tropaq~',ult.

In other words, when does or shouid an isomesic or warmer iso~emperature regime become of equal h~aportance at the grea~ i;rou~ level to soil properties? I would much rather see the use of tropic as a subgroup ~odifier .

Another alternaP.ve w'ould be ~o drop the use of the tropic modifier since the soil is already identified by the isomesic, isothermic, or isoh',¢pothermic ~emp'erature designation.

_Gu~. Smith:

This is not one ques,~io~ but a least three and I must answer one at a time. Lie asks, for ex~tnpl~, if an Aquult with. ~.n i~ohypothermic temperature regime has plinthite, fragipan, or ~ ; rupt textural chang,*.s cr pale clay cu rv~ it falls into z p~rtic~]ar grea,' group regardless o r the te~nperature regime. I should like f~.~.: ~o con--~aent on th~s one. The Aquults w~th plinthi~e are not uncommon in tropical re~onr~ whh isotemperature regimes; they become rare in the mid- l~titudes where the tempet'atnre regime~ are. hyperthermic or the.rm~c. They are unknown where

- 1 ( . -

Leamy Interview

the temperature is mesic. If the temperature is frigid, they are excluded by the definition of Ultiso!s. Though we have ;.he range of ~he Aquults with plinthite from the mid-latitudes where the temperatures are thermic or ~'armer and in the mid-tropics. Actually, we do not have any Plinthaquuits in the United States. to my knowledge. So that in devising the definitions it was not necessary to exclude so;.!~ wilh hyperthermic or thermic regimes from the definitions.

Fragipans a~-e virtually unknown in inte~ tropical regions. They are restrlcted to soils with thermic in Ultisols or mesic temperature regimes and it seemed unnecessary to write restrictions in the definitions that elimin;~ted soils that were not known to occur.

In the mid-latitude~, the soils with hyperthermic, thermic, and mesic temperature regimes are divided at the great group level according to the nature of the epipedon. We have the Ur:abraquults and the Ochr~,quults, ~n the mid-latitudes. If we d~d not write a requirement of temperature in the TropaquolL% we would have to divide the tropical soils according to Umbraquults and Ochraquaits. In the tropics, the amount of organic matter is very poorly related to the darkness of the epipedon. Therefore, having for the United States a separation of Aquults into Umbraquults :,ad OchraquMts, w? wouid have to carry ~his distinction into the inter tropical regions, where the color has virtually no relation to the organic matter content. ". herefore, we have brought in the te~nperature linfits ir.,to the Tropaquults, so that we would not be bound by the distinctions of Umbraquults and Ochraquults in soils where color has no relation to organic matter.

The second qi~estion impiies '~ha," soil temperature L~ not a soil property because it states that an isomesic or warmer isotemperature regime ~houid be of equal importance to soil properties. It is a basic ass~,(mption in Soil Ta::onomy that .soil temperature and sail moisture regimes are soil properties, altho'agh it is-d-~puted by many. The questioner is one of those who seems to di~;pute this a s sumpt ion . . i f so, he sho, uld, I think, devise hLs own classification system instead of complaining about Soil Taxonomy. There is ~,,) question but that soil temperature can be estimated approximately from altitude and latitude but there are still considerable differences between soils w:.th the same eleeation and latitude depending on aspect and also deoendin~ on climatic factors. We have for exampi¢ inter tropical soils tha~ do no~ have isotemper~cure regimes particularly in southeastern Asia. These may occur in other pa~ts of the world but the soil temperatures s~ .far have rarely beer. measured.

t|,t.' question prop.oses the 'as,' of the tropic modifier at the subgroup The third part of ' level or to drop it entirely as the soil is already ideatifiea by the soil temperature regime r~t the family tev¢!. This identifications, at the family lev~! is adequate where one has large scale maps and the map units are defined :n terros of family or series characteristics. In small scale maps where the ~aappiug units are mostly i~ terms of categoric units higher than the family, the soil temperature re3ime is not ind;eated b)" the name. In general, one may assume that the soils between the tropics have isotemperatu~.?, regimes but one cannot be safe in assuming that the soil temperature is isomesic or wanner between the tropics. It is possible that it is isofrigid, i f this is not indicated by the name of ihe great group, then the temperature must be inferred from a map that shows e~.-~va~'.ions of the soils, and -:be maps of the elevation are difficult for the pedologist to obtain but essen~ial for any interpretation~ whatever. The soil ',emperature regime should be i n d i c t e d roughly by the name if we are going to make interpretations c,f soil map units that are made on ~;ma|l scak: maps. If it is not implied by the name of the mapping unit, then the implication must be added as a phase and this complicates legends of small scale maps. There are already so many phase.*. -'.hat must be ~hown, slope, stoniness, depth of soil, textures, etc. It simplifie~ the matter of phases if the temperature and ~he moisture regimes can ~,¢ indicated by the name of the ,~axo~ that is used to identify the mapping unit.

A common feature of the three questions is the use of ~ropie. at the great group or the subgroup level. It probably is not ~ateri~l whether one uses the tropo m ~ i f i e r at the great group or the subgroup lev,.~ ~ e r than the problem that requires the ex~,ension of the umbric ep~.'~, on or the ochric ep';pedon importance into inter ~rop;_cal regions. The basic reason for using it at the great group level wa's T~o avoid the extension o f these concepLs that are applicable in temperate regions to inter ~ropk~d regions.

- 1 7 -

Leamy h~terview

Question 18

Further to that question, I would like to ask you why the "trop" formative element is used in sozne orders at s;aborder lev,81 as in Tropepts rather than at great group level.

Guy Smith:

In several of the O~rders, as in Alfisols, Ultisols, Mollisois, etc., the sut-ordera were defined l~rimanly by the soil moisture regimes. In these orders then, the temperature regime was brought in to be used on small scale maps as a subdivision t,f soil; with a particular moisture regime. Therefore, in :he AIf~sols we had Udalfs, Ustalfs and Xerallrs at the suborder level and at the great group level w~2 w¢.-e ~ble to recognize the tropic great group.~ tc avoid the distinctions according te the darkw~'.ss :~ the epipedon, in the Inceptisols the suborders were not defined on the basis of so~l moisture. Instead, we had the suborders of Andepts, Ochrepts, Ombrepts, etc., and therefore, the suborders of Tropepts was set up ;o avoid the distinction at the suborder level of umbri¢ and oc hric epipedon. Thir. is the same problem we had with the Ultisols wl,.ere we needed to ~,void .t~is distinction between umbric and ochric epipedons in inter tropical regions. It is possible that we made a serious mistake in subdividing the Inceptisols at the suborder level into Umbrept~ and Ochrepts. This is a problem that must be considered by another generation that has more experience with inter tropical soils thar~ was available to us when we were developing .";oil Taxot~omy.

Question lg

Leading on from tht~ last question, it is interesting to note that in the Andisol proposal we have separated Tropands at suborder level. Could you lsive the reasons for this'?.

Guy Smith:

The recognition of i_sotempe~ure regimes in the Andisols at the suborder leve! is parallel to the recognition of these temperatar~ regimes in the lnceptisols. The reasons are the same. In the Andisols as well as in the ]ncep~isols, Ultisols, etc., the color value is very poorly related to the content of organic carbon. If =F.e soils are no c separated abov~ the great group level where w e recognize amongst Andi:;ols, the Melanudands which are very dark colored, this color value must be extended to the inter t, ropi~l soils where the color value is not related to the organic matter content. On the island of SL Vincent in the West Indies, for exam,S.e, the northern half of the island is covered by a bl~.ek cinder deposit whi :h dominates the color of the soil The blackness is not related to lhe organic carbon, and the mer~ wea:?;ered the. cinders become the lighter they become in color. We war, ted to avoid using color value as an ind:~=ation of organic matter contents in the inter tropical regions.

- 1 8 -

Leamy lnterv;ew

Que :;tion 20

The next question is from Thailand and asks in essence, why the criteria used to distinguish Tropaquults are not used to distinguish Tropaquepts or Tropaqualfs.

Guy Smith:

The question is not quite properly phrased. In the Paleaquults ~,~.d Tropaquults, the requirement for color is only that the hue by 2.5~ or 5Y accompanied by mottles due to segregation of iron, or, if the hue is 10YR or redder, then the low chromas are required. Working in Venezuela, I examined the evidences of wetness." for aquie great groups and suborders, and made the proposal that the definition used for Ultisols be extended in all orders to the inter tropical regions. Namely, t!'e Inceptisols, the Mollisols, the Oxisols, the Entisols, etc., in Venezuela if they were wet, commonly marked the wetness by the yellow hues accompapJed by mottles. The criteria used for the Ultisols might have been applied more generally in Soil Taxonorn.y had we had few examples of other kinds of inter tropical soils.

Question 21

Questions regarding the spodie horizon and Spodosol, generally. The first one is number 8 in those listed by Professor Rust and reads as follows: the spodic horizon as oresent;y defined, relies heavily on laboratory data and in fact, can hardly oe established withot, t it. Does not this requirement put a considerable burden on the field soils scientists in their attempt to classify soils?

Guy Smith:

Soil Taxonomy specific~ly provides criteria for identification of spodic horizons in the field without laboratory data. In my experience it is rather rare ~hat laboratory data are required except in transitional soils where the spodic horizon is marginal to a cambic horizon. I have been criticized by the people working in laboratories that identification of tbe spodic horizon is too often by fiel,-I criteria and that the chemical properties are inadequately emphasized.

For f i e ld identification the only eouipment needed, is a rather powerful hand lens or a pocket microscope that is capable of giving the 60 power magnification. If with this lens the individual is able to identify the crack coatings on sand grains or the pellets, the identification can be made in the field without any laboratory analysis. It is also true that the spodic horizon reacts to fluoride and the Field's test for a',lophane is used in a number of countries wh*~re there is no volcanic ash, to ider~tify the spodic horizon and disti~guish it from the cambic horizon.

k

1 9 -

Leamy Interview

Qc, estion 22

We have ttad some trouble in New Zealand in eiassifS, i~g soils with spodic characteristics which have also high clay contents; these are some of the classic Kaur,i Podzols of the previous classification and we find that the horizon with spodic characteristics., may have clay conte'lts as high as 80 percent and if we calculate the pyrophosphate iron and aluminum to clay ratio on the basis of the clay content of the whole horizon, we cannot have a spodic horizon because it is far less than is required by the definitior... Can you advise or recommend what we should do about the classification of such soils which obviously have spodic characteristics to some degree?

Guy Smith:

I have seen such soils in New Zealand where the Kauri has produced, what appears to be a spodic horizon and I have aiso seen such soils in Europe where there has been a vegetation of Caluna. In both eases t.he high clay content appears to be due to the presence of an argi!lic horizon and the clay skins or coatings are commonly very obvious in the fine textured horizons. The spodic characteristics occur as a very thin layer above the argillic horizon and as rather thick coatings within the argillic horizon where there is obviously some sort of tonguing due to the illuviation of the clay. ~t is more or less reminiscent of a glossie horizon, but the zones in which the clay has been removed are now in filled with a very highly carbonaceous, very black material, that h~.s all the characteristics of the spodic horizon. If the soil horizon is s~,'qr01ed as a bulk sample the clay content does prevent the identification in t;ie laboratory of the horizon as a spodic horizon. However, if the soil is sampled not as a bulk horizon but as parts, then the spodic parts will not be found to have the high clay content; the high clay content is not the black material that you find between the pods.

Question 23

I

We have a small point of inquiry about the criterion for the spodic horizon concerning the ratio of pyrophosphate iron and alum;mum to percent c~ay, which is equal to or greater than 0.2. Ragg, Mckeague, and Simonson, have stz~xed that if the ratio is equal to or greater thr, n 0.15 it qualifies for a spodie (that ~z. using only one significant figure). Is this what was intended?

G_Q.qE Smith:

The intent of the definition of lh(" spodic horizon as well as in all definitions t',roughout Soil Ta:.:onorny that the numbers fo~1ow the normal mathematical rules for rounding. If only one decimal is used in the definitions, the numbers that are intermediate are rounded according to normal mathematical rules, e.g., 0.16 is rounded to 0.2. This was done throughout Soil Taxonomy except in the defin.ition oF t~:,e Andepts where one decimal too many was inserted and we used the numbers .85 for bulk density which is more precise than can be measured in the labora*,ory and we should ~ave used 0.9 or 0.8 or some number withoul two significant decimals.

::i:i~:i~iii ̧ : ~-

- 2 0 -

!

Queslion 24

Leamy Inl~erview

One of the greagest cha~ages in the clarif icat ion of Svodo~oii~ i~ New Zealan~ when we use taxonomy is tha~, previoc-.sly these soH:~ were raosltly c'alled ~odzoh~ arid w~rt: identified most readily ~, ~l,,e fiel~ by the pr,g.~ence of vale colored ~oriz~)t~s analogous to 'the albic l:orizon. Now that we have used taxonomy,, ;.his horizon has no impact on the classification until below family level. Some pedologistz are concerned at losing the utility of this horizon in their field identification and mapping and th~s question refers, in general tet,~as, why an albic ho,-izon in Spodosols is not given gre~er l~rominence at a higher level in the classificat[;~)n and in particular why there could not be a subo--let o f Albods.

Guy Smith:

On page 8 of Soil Ta.xono.my the sixth attribute that we des:red fo,- the taxonomy was that the differentiae should kee,o an undis'.-_urbed soil and its cultivated cr otherwise man-modified equivalence in the same taxon '~nsof~ as Oassible. If" the albic horizon is thin, the m~.~e clearing of the forest, seeding of gras.% and r, asturing car, destroy a rather --'- respec,,,..,le albic horizon. This I demonstrated in one of the type locations of one soil in New Zealand where, in the road bank there waz a good albic horizon bu~: if one crossed :he fence into the pas,~ure it was gone. This would mean then th~'t [r we ave. going to emphasize th~ presence or ubse::ce of an albic horizon more than the presence of a spodic horizon, one would have to draw a boundary along the fence because that is where ~he ~.ibic horizon stopped. It would be possible of course to emphasize the albic horizon at the expense of the n~,ture of the spgdi,~ horizo~ and if we did that, we would have perhaps an Alix~ and a Chromo~ and then these would l~e subdivided at the great group level into humic and other types of spodic horizons. We i'eit when we develop taxonomy, perhaps erroneously, th~,t the nature of the materials that accumulated should be g:ven greater weight than the presence or absence of an albic horizon. Certainly if one were to emphasize the importance of ~,h~ albic horizon, the de?inition would have to require that it extends .'.o depths greater than 25 cm; otherwise plowing would change the nature of the classification of the soil.

Question 25

In many young landscapes (this is particularly so in the south island of New Zealand), the landscape comprises Incep~isols gr~0ing to Spedosols and detaile~ soil surve~.s in this environment reveal intergrades betwee[~ l~ceptiso!s and Spodosols. There is no provision in Soil Taxo;iomy for spodic subgl~oups of l~.~eptisols. Cou!d you commen~ on what should be done in a s-~tuation like this?

Guy Smifll=

- j

The Soil Taxonomy was d~igned so that the ieas~, poss:ble disturbance would be made if new knowledge and experience indic~!,,ted that we should change some part of the system. In this sitgafion where an intergrade may be desired between an Inceptisol (a Dys.~rochrept) and a Spodosol, the people who hRve some experience with the,z soils must propose that this intergrade be introduced into the sys~m. In making such ~ proposal it would be essential that the man who m a k ~ the propo~d: proposes also a definition for the spodic subgroup of the Inceptisol. This is perhaps the mo~t difficult part for making a proposal for a change. We need to have not only Zhe proposed definilion but .we need also to h~ve some reason why the change should be rhode. Does it improve accuracy of interpre~tio~s, if so this should be spelled out in the proposal.

-2,) -

Lea:~y Interview

Question 26

The ne~t question corc~es from both New Zealand and from Thailand and ce~.cerns lhe identifieatiot~ of base ~9.turation l,.~ distinguis~h between Alfi.so~,s and Ultisols. The Tha: question reads, why ~:, ~rercent base saturation dezerrnined at 1.25 m below th,' top of the argill~,: horizon or at 1.8 meter below the soil surface. In addition, the Thai pedologist would like to know what to do if at 1.8 meters, there Ls a iithologic discontinuity with contrasting material. The New Zealand question asks, what is meant by identificatio~ of base saturation at certain de0ths below the argillic horizon. S~¢onoly, it asks if the sample should be from a thin layer or i~s an auger sample from the a~,groxio~ate depth satisfactory .or should the whole horizon in which the depth occurs be. ~mpled . Thirdly, ~"h~-'t is the intent of the requirement of b~ze saturation at depth particularly when tiae ~ystro-eutro distinction is made on the basis of base satu~tion of the n~ajor part of the prof i le?

Gu__..~v _Smith--

The first cerumen! is, on wh:at is n~eant by the word at a depth of one and a quarter meters or 1.8 meters. This meanz according to the English language, "at'. It can be mea~t~red in one of two ways. Either one rakes e, ~ample of z thin .~ubhorizon at the specified dcpth or one samp!~ all horizons ~bove and below,." the critical depth and then makes a smooth curve of the data and the depth at wh~:h lhat curve crosses the 35 percent base saturation is either above or below t,'~e critical depth of 1.8 ra or 1.25 m. If the smooth curve cro~'.~es the 35 percent bas.'~ saturation limi~ at a depth z,~,allower '~,hzal the critical of 1.25 or !.8 m then the base saturati, on is cert,~inly less than 35 percent at the critical d,:. • th

The reason for tl~e ch~.~cc- t;,f 35 percent at the critica, ~ depth sl~ecii'ied, is the simple one :hat is common to all the def~nition.,~ in the taxonomy. it is that we got groupings ,'hat permitted us -, make more statements .and more precise statements about the soil use then we could otherwise make wi~h another ~Arait e~f base saturation or anothe, ~ limit of depth.

If there is a lithological discc, nzinuity at or above the critical depths of 1.25 or ! .8 m the b ~ e ~aturation at these cr.;~ica! depti~ ig still the 35 percent limit between the Alfisols and Ultisois. The base saturation of a s~eeif.lc horizon is not just a property of that ,~pecific horizon but it ret?,.ects the er,~tire proce.,,.~ of le~.chin,~g ~nd recycling of ba.,~es in the soii which affects the ,,~hol¢ soil in all horizons not just the one ~:orizon. The base saturation curves are quite interesting propecties of the whG.ie soi~ rather than of any specific horizon.

Question 27

k

Qu~tion number 3 compiled by Professor Rust reads why was base saturation chosen as the (only?) differentiating criteri;~ between Alfisois and Ultisols? B~ using it :..,t different levels, does not this "violate" some rule of !ogle and or taxo,,~omy?

Smith:

As we pointed out in the 7th Approximation (Soil Survey S!aff, 1960), taxonomies are devices of men made for sgeeifie prarpe~es, not truths that we have discovered. This is one gene~ l ruler of logic and of ¢i~sification. It has been recogr, ized in the b~ological taxonomies

.starting with Linnaeus. In the Linnaean ~onc~my of plants and animals, the prineipa'., kind of plant o~ animal, was the species. Linnaeus said that a botanist must know and remember every gonus~ (Cain, ). John Stew~.rt Mill (!$91) pointed out, that objects had to be classified for a

k ' L ~ ¸

- 2 2 -

Leamy interview

: ? - ' , ,

purpose and ;.hat if :here were different purposes ~rhere could be several classifications or t~o,',,om~e~; t'.~e called them scientific clagsffications, ,:~f t~e same objects de0ending on the purpose, l~!iil ga~d that the b~ t cIassification was )he one that permitted the largest number of t;,,e most ~m~'~or'an, r statement about a given class of objects.

There is a distinction betwee.,~ the taxonomies of plants and animals on the one hand and so i l on the other. The taxot~omies of livin~ organisms have in the past generally been built on the phylogenetic prh:ciples, namely that of descent. When we try to classify soils, then there are no principles of descent. There are no common ancestors of soils. They are not living org~isms; they are as we ~l! know, in the border line area between the biological and earth :-ciences. But the logical principles of John Stewart Mil!, that the best taxonomy or scientific classification is the one that permits the greatest number of" the most important statements about the objects that have been ci~ss~fied still applies. In soil science these important statements are our interpretations., not our theory of genesis. Therefore in order to be able ~o make any statements of any sort about the classification or taxonomy of soils it is necessary to use the same criteria at different ~tegorie~. The biological taxonomist m~tkes very little, if any, use of tt~e orders in the classes them~elvej; he is concerned w;th the species. The pedologists who must make soil surveys at vary:~p,g sc, zle~, makes principal use of the higher categories with smah scale maps and principal use of the lower categories w~th large scale maps. In order to be able to make any 3tatementz whatever about: ,he different orders, that can be used with e~tremely small scale maps, the most importa~;t characteristics to the use of the soils must be used at the order level. With somewhat larger ~raa~,| scale maps, the map units m~y be defined in terms of subord,.-rs or great groups (always. in ~erms of phases of course); we may want to make somewhat sma!ler distinctions in the characteristics that we have used between )he orders. When we get to the level of gre~t groups or subgroups used in somewhat iarge~ scale maps, the same characteristics that we have ~med at the order level may or may not be important but we use them when they are importa_,',~t, important i;~ the sense that they permit us to make statements about the clasges th:~t we m'e showing on our maps. in the detailed soil maps we must have all of the accumulated differences to the maximum extent possible, of features or properties such as base ~ tora6on , ~oit moisture, soil temperatt~rc, many, many properties are used throughout the taxonomy. The general rule i~ that we make the least subdivisions c~f prol~rties at the highest categories and tt~e maximum subdivi~ic.ns of properties in the lowest categories because the lowegt ~ e g o r i e s are only used in the large scale maps. The biologists who do not make maps at ,Jifferent g~ales are noI con.cerned wieh this particular problem. The problem Ls that cited by Joht~ Stewart Mill., that the be~t taxo;to~,~D, is that one lhat permits one to n:~.ake the most statements a.t~out the most important properties of the units that have been claszified, t't is one of the strengths of Soil T.,,xonomv not the weaknesses iha t we use the same properttcs in different categories. The .¢eewest subdi~,i~;ions of a given property are u~ed in the higher categories. The largest h u m o r of subdivisions of the property are used in the lowest categories. This is completely i~ agreement with the logic of John Stewart Mill or any other taxonomist I know. If the pedo/,ogist were to read ',he modern literature of 'the taxonomy of pla~,ts or animals, just a few books, he would find statements such like the one I would cite from C.ain, that if the botanist or zoologist were f~ced with the problems of cl:~ssifying plants or animals according to variability over both time and space, he would find the present syztem intolerable. The variabi!ity c,~'er sv'~-~' of a s~ecies of animal; is somezhing that can be observed today. The variability over time ~s something that depends on the fossil record and is not only imperfect but often complem|v t',bsent. In soils we must deal with this variability over time and over space. This is therefore a major difference between taxor~om~, of soils and taxonomy of Eying organisms. We therefore must deal with the taxenomy of soils in a somewhat different manner that do the botanists or zoologists deal with their taxonomies. Actually, Cain has sugg~.sted that it would be better to drop the system completely and devise a new system but be says because of the priority of nomenclature, we would get into so many trog~bies that :.t i:; probabi, y not worthwhile. When we started to devslop Soil T~'conomy we decided that if we were goir~g to make a break with t he past we had best make it completely at this moment that w~ publish Soil Taxonomy.

The reason that we cho~e _b~e .,~aturation to separate Alfisols and Ultisols w~,,s again that it let us make the most precise state, merits about the soils included in Alf'sols versus Ultisols. T h e i~arfioular~statements that w e ~:anted to make about these two classes of soils depended on whether they could be used with or without amendments. The Ui,,Jsols cannot be used for a

= 2 3 -

Leamy Interview

permanent agriculture without amendment.s except throug,~ a system of fallow. "rh~r -'%-e,~.,ots'- have been used without ferti l ize~ or amendments for some few ~.housands of yczrs and a permanent agriculture can be based on ".~e A|fisols without amendme~:,,t. Th;..s ia not generally ,z~ with Ultisols, which aftvr a time, gene~a"y the soils become sc~ depleted of )",ases that they cannot be used for agriculture at all even with fal)ow. This seemed an important statement that we would like to make about the order, at the order level, and it was therefore the base saturation that we selected to rc'flect the difference l'etween the soils requiring fallow periods and soils that could be used permanently.

Question 28

I have always been interested, i.n the rhodic great groups and in some of the rhodic subgroups because of the ~" ," . ;, uzs,,mcL ~e color and I have often wondered what this color signifies in terms of accompanying chai'ac~eristi~s a~c~ what was the reason for setting up the rhodic great groups.

Guy Smith:

If o~e ex',mines the so[L~ tha~ are gncluded in Alfisols and Ultisols at any given suborder level, one finds generally that there are s~qls tha: have severe problems of soil structure. The physicists, the agricultural e, ngiaeers have not been able to quantify methods for measuring soil structure and yet this is one of the most important properties to agriculture that a soil possesses. Examining the soils that are Cl~L~sified as Xeralfs or Udu!ts, one runs into soils that have no structural problems occasionally bu,, generally there are serious problems of soil structure under cultivation. The soils that do no~ have ~erious problems of structure, we first had found in the U: S., always had dark red colors. The original definitions of the rhodic great groups required value~ moist of 3 or less and hues redder ~i~an 5YR. In Tazmania, I was shown soils cn .a single lava flow from hie h -~--,,~:,,.~ ,.,c,,,.,,;a to se~. level, and the hue became less red as the elevation increased but the ~tructurai stabifity of the soils was the same irrespective of hue. The value did apt change with elewation. Therefore, in the Ultisols in defining the rhodic great groups, the requirement of 1he red hue was dropped but the requirement of the low value was retained.

Because, so far as we now know these soils are always developed from basic parent m~.terials such as basalts, limestoner,: etc. The contents of phosphorus are generally higher in the rhodic great groups than in the other.,. The use of the color value and the chroma was predicated on the assumption that these features were correlated with the structural problems, with the phosphor-us conte~s, and so ~n. There were many cow,~rying properties that were extremely important to soil use in the rh i : ~q._ great groups. No matter where one finds them they ar~ about the most intez~ivcly farmed soils of the particular suborder. Rhodic great groups were not set up in Moi~isols becat,.se t~,~e:e were no particular differences in soil structure with soils th=~t have a mollie ep,;pedon. The formative elemenz 'rhodic' implies red, whereas the actual characteristic used in Ultisols is the color value. This may disturb some people but cme must recall that there are rhododendrons that are purple in color.

r - 2 4 -

Learr:y lnte~ view

Question 29

Some soils in the tropics and in Thailand, in particular, have a thin argillic horizon over shale. Because the shale is b.igh i~ -.:lay, the clay percentage does nc~t decrease with depth. The question is, should these soil~ be classified, as 'pale' great groups.

Guy Smith:

There are pale great groups in several orders, Alfisols, Aridisols, Molliso!s, Ultisols are examples, and in each order or suborder the definitions of the pale great groups vary. In the Ultisols, the pale great groups mast in addition to the clay distribution lack very many weatherable minerals, la the other groups the definition varies, suborder by suborder but have no relaOoa to the clay distribution alone. In every definition there are characteristics other than the clay distribution and in add~.tion to the clay distribution. In some pale great groups, an abrupt textural change between A a~d B is used as a part of the definition of the pale great groups. In others a redclish hue or mottlez of high chrt, ma are a part of the definition. It would not be. intended to group a s~il in a pale great group solely on the cla7 distribution.

Question 30

In Soil Taxonomy, a common strategy is to chose a pararaeter which has morphological expression and is covariant with o~.her properties which are of interest for the classification. A case in point is the use of the presence of secondary carbonates to distinguish ustic and udic subgroups or moisture reg:'-~es. "I'h~ parameter was chosen because it was assumed to have covariant properties wherever it occurred. However, testing has indicated that it is no~ always covariant and in many iandsca~-.~;s, particularly in non continental landscapes, where loess is nonca!careous, moisture regime changes are not recognized or not accompanied by secondary carbonate depositions. Would you iike to comment on that in particular, and also on the flexibility that Soil Taxonomy ha~, to accommodate to that situation?

Guy Smith:

' S

It mt~st be remembered that, while Soil Taxonomy was intended to group the soils of the United State,,j, with which w,~ had experience, it was also intended that it should be possible to exlend the definitions so that they would be applicable to soils of other countries. In the i lnited States the soils w~th u~tie or xeric moisture regimes are almost always from parent m~terials that have carbonates or there ~re carbonates in the dust that falls on the soils. The original taxonomy used in the Uai~ed States, that of Mzrbut, divided a!l the soils at the highest category according to the r~rezenee or absence of a horizon of accumulation of calcium carbonate. "l'he emphasis on t.k~ hoxizon has been greatly reduced in the classification of 1938 and in Soil Taxonomy. However, the prejudice in favor of using this horizon continues to exist because of its long traditional ~ e in cl~sification. The definitions ;.~ which the presence at a given depth according to parficie size distribution of a horizon of calcium carbonate accumulation, assumed a relationship between the depth of water penetrztioa into the soil which in turn.was correlated with the moisture regime. The limit~ of depth were selected according to the traditional concopts that ~:~e depth ~o the earbonazes varied wi~h the rainfall. These were alw.~y~ in regions .in which the rainfa!l was limited, and genetically the depth to the horizon of aeeumu!~tion of ca l c ium e~x.bonz~e was a function of the total rainfall and of the soil temperatu:m.

2 , 5 -

Learoy Interview

In Venezuela, ~ found that the soils wi,,h ustic moisture regimes and with dry periodv ranging from 6 to 9 months had carbonate accumulation at depth provided " , irma, the ;)~rent materials r.vere calcareous. Noncalcareous parent materials gace rise to soils without carbonate accumulation irrespective of fae leng.~h of the dry ~eason, that is the length of time on the avecage during which the moisture control section was partly or entirely dry. Therefore, we had soils from noncalcareous ma',eria~s that were marginal to Aridisols but had to be placed in udic subgroups by the definitions in Soii Taxonomy. This is irrational that correlation between the depth to carbonates and ~he moisture regime is very imperfect. The relationship depends not only on the amount of re, infzll b~t en the distribution of the rainfall and on the carbonate content of the parent mateHal~. Therefore, in Venezuela, havin8 reviewed the application of the definitions of Soil Trzxonomy to aoii~ in a wet/dry tropical climate, it wax obvious that we could not use carbonate.,.., as a basis fer defining udic and aridic subgroups of Moilisols er Alfisol.~. I therefore proposed that ~he definitions be changed and that the depth to secondary carbonates be eliminated completely from the definitions and that the definitions be rewritten on the basis of the length of time during?, zne average year or during some percentage of years ~hat the moisture control section wr, s dry in some part or in all parts.

Question 31

The next question is #30 in those compiled by Professor Rust and asks why was the argillic horizon clay increase ~' "" 20 overlying eluvial horizon? e..~al:hshed as a percent relative increase compared to the

Smith:

The French Taxonomy uses ~n increase of 40 percent as a basis for recognizing different classes, particularfy the so! lessive. Amongst the Mollisols, the existence of an argillic h~')rizon is rather widespread and marks the br:ak between the late Pleistocene Mollisols and the Holocene MollisoLs. In these soils the break between ~he eluviai and iiluvial horizons is at o.bout an increase of 20 percent in clay. Tb~ is actually the minimum limit in the Mollisols at which we thought the field man coulc~ identify t.'.qe change in the particle size distribution or texture. Therefore, in Soil Taxonomy we took the absolute relative increase of 20 percen~ from the Mo~iiso1~ as our minimum for recognition of an argillic horizon. In Alfisolr and Ultisols, the normal situation is that the ~.ncre=se is 40 percent or more. it must be ~'emembered that th~s increase of 20 percent is applied only _to soils having clay conte~ts ranging between 20 and 40 percent in the eluvial horizon. The 20 percent increase in a soil which has 20 percera ,:lay means the field man must d:~t~nggui.~h ~ t w e e n 20 and 25 percer, t clay. with his fingers. We desired to have d e f i u ' t i o ~ that could be applied in the field without referring '~amples to the laboratory.

Question 32

The next question is #31 in ~hose compiled by Professor Rust. Where does the criterion of 1 percent or more of oriented clay in the argillic co,he from? We think it is very difficult to follow the criteria with reasonable precision because the statistics of poinl eounfng in the

- 2 6 -

Le,qm y Interview

n(;crosco~e, show that we need .qbout 3,500 com:*..s f6r each thin section to be 95 percent sure in a 10 percent error.

Smit._.__hh:

One mast first remember that the limit is 1 percent of the oriented clays not 1.0. A 10 percent error therefore is no: o)~y permissible but probably expected. If the point count shows 1/2 percent or more, the rounding of r.umbers w,-'U bring it to the necessary 1 percent.

The assumption was, in this limit :hat if the clay-skins could be identified in the field there would be a leagt I pei~cent in thin sectie,s. This assur, ption may not have been justified and it was proposed for testing bat t~at no one had ever criticized it when it came ~.ime to print Soil T~xonomy.

Question 33

The next question, I would like to ask is~ I believe i~ is a widely held assumlotion of practicing pedologists anyway, that both th~ presence of clay skins and a clay bulge are required for an ar?illic horizon in most cases. ! understand now from talking with you that the emphasis on this is not ouite correct.

Smith__

Surely, no clay ~crease can be required in soils that have beez true.outed o,r in soils in which there is a lithologic disconti~iuity giving rise to an argillic horizon even witll !ess clay then in the surface mantle. In these cases where there is ~o increase in clay between the argillic horizon and any overlying horizon, such as a plowlayer, we do require 1 percent cutans. This lim;t of 1 b'~.,rcent was set so low that if the pedologist in the fiold would ident;fy some he surely would have the necessary number of clay skins in the soil.

If the clays have 2:1 lattice minerals, the argillic horizon does not need to have clay skinu, i f the.re are skeletons in an ov~r!ying horizon. It is pointed out in the discussion, rather than the summary, that in som~ places Varticularly those with wet/dry seasons, the ~ " , .~ec~a~ field studies are needed more tha~: laboratory studies to identify the presence of the horizon with illuvi~2 clay. If the polypedon has a range in elevation and if the boundary between the surface ~.nd the heavier textured undo:flying h o r ~ n is clear or abrupt, it may he necessary ~o trace ~he finer textured horizon ~aterally ~'o be sure thzt it is not a depositional feat'._'re. If the increase in clay is marked enough to be o~servable, and if the boundary iJ c~ar or abrupt it is extremely difficult to assign the origin of the finer textured horizon to differential weathering. Moisture conditions i~ the soil d6 not change abruptly but rather gradually. But in the discussion rather than in the summary it is pointed cut tha~ the field studies alone can realistically indicate the i!luvi~l nature of the finer textured horizon.

It also pointed ou~ that t.he significance of the argillic h-.,rizon to soil genesis is not partic~'.tarly more important than that of any other kind of diagnostic horizon. "lbo much attention generally has been g~ven tc~ the presence or absence of clay skins in soils. The important thing about the clay skins is that normally they have marked inf!uenee on the amount of nutrient elements that are cycled by plants. They have more nitrogen, phosghorus, ~ tass ium, than do the ped i~eriors. If the finer textured subsurface horizon is not actually illuvial it is not so important to th~ plants as is the nature of the nutrient contem of the pod e.~.atings of the sure'aces of tide peds. "This is t~e important thing in relation ~o plant nutrition. In many soiis with extremely low fer t i l i~; soils in which the nutrienis are maintained in the soil by p~an.~ cycling, the roots are able ~o reach the subsurface horizon and extract water because

- 2 7 -

Lea(ny lntervi~w

calcium i~ cycled and i.~ pre~ent in the coatings on the peds; voids with coatings along which water carrying the recycled nutri~mts n~.ove. Without the calcium the rocks cmmot enter the subsurface horizons attd therefore cannot utilize the available water and the soils become extremely droughty,

Question 34

The next queV, ion is re'tuber 32, compiled by Professor Rust, and it asks, what is the reasoning for using the illuv[al horizcn (or sL~rface) as a reference for the required increase in clay ~'or the argillic horizon, instead! of the underlying horizon? In other tyl~.s of pedogenic horizons, such as ca!cic, gy~sic etc. the C horizon is used as, at least partially, as a reference point. It seems 2o me that in soils tha t are truncated, the C horizon would be an acceptable reference point. Likewise, the C horizo'a could be a reasonable reference point in a soil developed in a relatively uniform p:r~rent material in a stable landscape.

Smith:

There are several difficultk:s in u:;ing the C horizon as a reference point for identification of an argillic horizon. One hi that, we would necessarily ha:,e to u~e the statement that "the clay decrease in an underlying horizon" because we have consistently avoided the use of A, B, and C in Soil Taxonomy. The ped¢~16gists can argue endlessly about what is B, and what is C without ever reac~ing an agreement a~d the base of the argillic horizon, as assigned by different ~edologist, will then vary 8reatlit in depths. Secondly if there is a lithologic discontinuity nero the b ~ : of the argillic horizon, then using the underlying horizon in such situations means that the definition cannot be applied uni~,ersa!.';y or can only apply to some kinds of soil where the parent material is uniform. What th6n does one do when the parent materials ate not tmiform. In this slmation one must have so~:ie other kind of reference and the only universally applicable differential ~:hat I can find wa2 the presence of clay skins in the horizon that we would like to call an argiLlic horizon.

Question 35

The next question is number 28 from those compiled by Professor Rust and asks how did you arrive ~t the,* general concept of diagnostics horizons.

Guy Smith:

In the early approximation that led up ~,o the development of Soil T~onomy we tested the groupings of soils according ~o th~ nature of the horizons; soils with A hor izo~ only, soils with, A and B horizx)m where grouped into separate taxa in the appr ~ximations. We shortJy realized Hat the nature of ~ B horizon was ,;mg:.'ortant and we began., to talk about the t ex tua l B horizons the podzol B horLzons, etc. Th~s was the first step toward the use of the. diagnostic ho~zams. It was not too many approximatio._= ~!ong, before we re~.ized that ~he pedologist would not agree on a B !~orizon. And yet this has been used in !he highest categories of the appro.vAmations. I sent ow, the list of ~pproxim;itions in whici~ ! s[~ ~ke of a latosolic B horizon and g a v e a partial definition of what it should be like. The comme;ats I received are mostly

, L e;~my In:ervlew !

concerned with whether or not if this was a B horizon and I ~,ot no comments of whether the definitior, will produce useful g.roupings or not. So at this poD~,t I :,~opped referring to A, B, and C horizons and started the ~se of the diagnostic horizon~.

There were further problems that some of the diagnostic horizons could, only with great difficulty, be considered A, B, or C; there were other kinds of' horizons. For example what is a duripan? It is not an A, becmtse its a subsurface horizon, -'.'t is a horizon of acccmulation c,f silica and occasionally of iror~ but is it a B horizon. Pedologists generally would not consider it a B, horizon if there were an overlying arlgil~;.c horizon which is a very common situation. The Canadian pedologists com, ider the horizon of accumulation of carbonates as a B horizon; American pedologists cons;.d.er it as .a. C horizon, Cca versus the Canadian Bc~. and there seemed no prospect of any international agreement on whether a horizon of carbonate accumulat.;.on is a B horizon or a C horizon. Therefore the use of A, B, and C was impossible in a general system because of lack of agreement amongst pedologists and, the only al!c:rnative was the ..'.ubstitution of diagnostics horizons about which *he original concept of A.., B, and C would not interfere with general agreement.

Question 36

The next question is number 53 from those compiled by Professor Rust and asks g, hy is 12 i~ilograms of organic carbon used to separate humic great groups.

Smith:

There is only one greet g~'oup, Humitropept. where 12 kilograms is ~ e d as a limit to separate the great group from others in the sam~' suborders. There are some suborners, ~ in the case of Humult, where 12 kilograms of carbon iS used. T~:zre are other humic suborders such as Humox in which the fimit is ,~,-'~*. 12 but 16 kilos of carbon. The reason is again the same as in other questions ~hat the grouping= that. were ach}eved by using ~hese limits seem to permit more statements about the taxa that were formed and more accurate interpretations. We tested limits of 20 kilos of carbon, 16 kilos of carbon, 12 kilos of carbon, and no one limit seemed to be useful in all of the different orders. The 12 kilos of carbon per cubic ~e ter seem tc give the best groupings for UltLsoL~ and Inceptisols but a higher limit of 16 kg seem best for Oxisols. In every case, the limit w ~ not strictly on carbon b~.'t also involved the temperature regime of the soil.

Question 37

In many cool temperate forested regions, litter accumulates to considerable depth bequeath the forest trees. The question ask,, how to distinguish between an O horizon and a Histosol

:when you have litter to depth of greater than 60 c~. It doesn't seem as if it is appropriate to classify such a soil as a Histosol hut i t is ~o t clear how you avoid it.

- 2 9 -

Leamy Interview

Smith:

You avoid it by changing the definitions in Soil Ta.ronorny. Nobody is perfect, and there seems :o be some confusion in the definition of organic and inorganic soils. I know that under th*. Kauri trees in New Zealand the O horizon can be more than a me~,r thick near the tree, getting a few meters away from tl'.e trunk of the ~ree, the litter becomeg much shallower. Nevertheless, one does not want tc have a complex of Histosols and mineral soils with the limit being ~. small circle around the trunk ,;f the existing tree. "I'~.:~ definit 'on in Soil Taxonomy', however, requires that the soil be classified as a Histosoi if the 0 horizon is more than 60 cm thick. ~,ome clarification in th'~ rmxt addition of So;l Taxonomy seems to be necessary and as a general rule, when one runs into a si~:uation of this sort where the soil is obviously mis- classified, some comments should be mzde to the Soil Co,~.~ervation Service so that they will be aware of the deficiencies in t he cu,'rent edition.

Question 38

The next question concerns the fragipans and asks spec;t-ica;iy why or what is the genetic reason for the condition in the descri;aion, that f~gipan nas an illuvial horizon. Comment is made that not many New Zea!and fr-~gipans have an illuvial horizon. The question further asks to comment on whether or not there ~hould be a fragic diagnostic horizon which would be one that does quite meet .'.he fragipan definition - nothing particularl:e with regard to root s0acing.

Guy Smith:

The reason for the s*.ztemen: that there is a illuviai horizon above a fragipan, unless the spit has be~n trun~ated, i.~ ~imply that, ;n the experience that I have had, such a horizon always ~xists. Because we do not u.,,der~tand very well the genesis of fragipan there is no genetic reason to this. It is only a matter of general observation. It is very likely, now, that I have had an opportunity to look at frag.ip~ns ~.~ some length in New Zealand that the iliuvial horizon that ! described is one in which ferroiy~is has been an important factor, ferrolysis being the destruction of clay under .~iterna*.ing wet ~'ad dry conditions. There are enough problems in identification of a fragipat', that we eventually decided in the U.S. that i~ was an either/or situation, that the soil had a l'~agipan or i" did not have a fragipan. But we did provide for intergrades where the britt.leness was obs.'rvable in an a1~preciable part of the fabric ot the horizon but the roots were .present at intevvals of less than i0 era. For exampte page 129 Soil Taxonomy, we provide for a Fr.agie G£ossudalf. They are like the typic except that they do not have a brittle matrix. It is much simpler to define these subgroups on the percentage of' the matrix that is brittle when moist, then it is to establish z new diagnostic horizon that is somewhere between a fragi,~an arm no fragipan.

Question 3g

. _ 2 .

i " ; ~ -

T h e next question comes from Thailand and coi~¢erns the particle size class at the family level and the question is why do w¢ ,r~ve to use the weighted average of the control s~ t ioa for soils containing coarse f~'agments (2 mm in diameter or larger). A thin and dense band of

- 3 0 -

Leamy ln~.e~view

gravel such ~ ironstone can make root penetration impossible even though the weighted average of the iron stone is less than 35%

Guy Smith:

It seems probable tha¢ whoever a~:ks this question is concerned with a diagnostic horizon that has not yet been defined or n~med. While I have seen many sculls in the tropical region that comain consider~bly more ~han 35% ironstone, I have never seen a horizon that was important to the penetration of root.~ or water. I have had proposals for a diagnostic horizon containing ironstone bu~ the propozals were not acceptable because they will transfer any stonelines in tropical soils into a diagnostic horizon. In other parts of the world, stonelines are not recognized as diagnostic propertie& Soil Taxonomy gives considerable weight to horizons that interfere with the penetration of roots. If such horizons exist in the tropical regions there should be some proposals f~r a new diagnostic horizor~. In my t~aveis, no one has ever concerned themselves with st~owing me s,~cb, a horizon, so I assume that it is not an important property. This is an as.~uraption and i~ r,~ay not be .~ .valid one for interpretations of soil on a particular plantation or ~ particular tract of land that someone wants to cultivate. Soil Taxonomy does not concern i tself with areal extent. We are concerned with the interpretations that must be made for a particular ~.r,~ct of land whether or not this a common or extensive situation or whether it i~ rare, The teg~nd for the soil map of the world by FAO and UNESCO, is one that of necessity depends on areal extent. Soils lh~t are very extensive in the world can be shown. Occasional soils o f small extent cannot be shown in their presence while important. A particular trzc¢ of land cannot be indicated ia the legend because in the world as a whole they are exceptional even though the properties are ext*emely important on that particular piece of l and . Theref,'~re if there are such horizons with ironstone that are impenetrable by roots someone would suggest a new diagnostic horizon just as I have, since retirement, suggested the densipart and the lithoplinthic horizons. There are certainly other horizons that should appear in taxrmomy that have not been suggested.

Question 40

I would like now to discus.-, wit.h you generally the use of climatic parameters in Soil Taxonomy and perhaps we can start by using one of ti~e questions compiled by Professor Rust (number 2 Jr, his list) whict~ ~mtes why were temperature and moisture chosen as differentiating criteria when in fact they would seem to be external factors to the soil.

Smith:

It would appear from ,.he question that if one ~nserted a thermometer into a soil one would not get a reading; the so;.l iias no temperature according to the question. This i~ a rather general ;~roblem with people who have not had an experience with soils over a wide geographic range. The temperature at one mo~r~ent or at one day is no, necessarily the s_nme as the temperature at another moment or another day. Yet there is a temperature. When the late Dr. Kellogg w e n t to Canada some years ago to examine the reasons why alfalfa (lucern) was suffering from def]cieneie~s of .~ulfu.~, he carried with trim a thermometer. The soils have a layer of gypsum at about 50 cm der.,th and yet ti'~e, alfalfa was suffering severely fcom sulfur deficiencies in Lhv presence of g~,,~sur.a. He demonstrated to his Canadian host that the horizon that con'r~_Aned gypsur.q had a temperature that was too low to permit the alfalfa roots to enter and he demonstrated also :that there were no roots in that horizon. Is this then not a soil property? In my judgement it is the low soil temperature that pr~.vented the alfalfa roots from entering the horizon with g.wdsurr, and obtaining the necessary sulfur. What causes the low temperature of the soil n'iaybe, the climate perhaps and probably .is but still it ~, also a soil property. The soil t e m ~ r a t u r e can be ~.ncreased in the summer by removing an insulating layer

- 3 1 -

Leamy Interview

such as an O horizon so that with a given climate the soil temperature is not necessarily the ~r, me in soils that are undergoing the ~liffe, rent uses. This does not mean however that there is no temperature. The soils of Northern Canada have very different temperatures flora the soils from the West Indies. The soil temperature is not only important to the growth of plants, if it becomes low enough to impede the growth of the roots and it is also an important cause of sol' differences. The temperature is exceedingly important in the rate of chemical processes and therefore in the rate of weatherit.,g of" the primary minerals of the soil pare~at material. It is a basic assumption L.7 Soil Taxonomy tha: the properties thr2 are the result of genesis or th..~:t are factors in the genesis and there/?,i'e ca.,~'es e f other properties, are the factor,~ thac should be used in the definitions. John Stewart Mi,~l pointed out ~hat properties that are causes of other properties ~.re preferable in developing a class;,fication.

Question 41

Soil temperature and mois,%zre parameters are not always used at the s~.mc categoric level in all orders and notably in Inceptisols and Entisols, moisture regimes are not used any higher than the great group level; in many of t:he other e rders moisture regimes are used at suborder level. Could you comment on the reason~ behind this?

Smith:

Consider the Entisols as an example. Entisols have nc~ diagnostic horizons other than an anthropic epipedon. One could have used moisture and temperature to define suborders of Entisols. Certainly this is possible b~at the question is ol~e of develo:~ing classes about which one can make the greatest number of stmements about the things included in a given class. Amongst the Entisols there a,:e s~;ve~l re~ons why the soils do not have diagnostic horizons. One is that they are continualiy ~-eceiving new sediments. Another is that erosion is removing ma:erials more rapidly than allows horiz~gns to develop. The third one is that man has disturbed ~he soil to great depths and mixed h oriz.on~, that have previousiy existed, if one considers then, these reasons why Entisols have no b, orizc, ns, it seems that one might be able to make more statements in common about the soils which are receiving the alluvium than about the soils which are alternately moist ~.nd d~2¢. Having decided to divide the Entisols according to the reasons why they lack horizons, although these are not specified in the definition, the next most important features of the s o~ls seems to be moisture and temperature. At the first category possible then, naoisture and temperature were recognized zs differentiae but in Entisol the suborder took up :he causes for the h~ck of horizons and therefore the introduction of moisture and temperature could only be made at the great group level. Had we insisted on using one criterion at the same categoric leve! uvder all combinations of other properties we would have had an almost infinite number of categories and we would have been unable to make many statements about most of the units that resulted.

J;.

Question 42

The next question is. how did you cotne to develop a concept of isotemperature and what do you think the significance of isotemper~ture is to the Soil Taxoncray.

- 3 2 -

. • - , -

- k

Leamy Interview

Gu~ Smith:

First, one must keep in mind that one of the purposes of developing Soil Taxonomy was to facilitate interpretations about soil ,ase. Consider the differencez between soils that ha,;e a mean annual temperatu~-e i>:~rhaps of 10 to 12 degrees, one soil being in a temperate region ano the other in an inter tropical region. The growing season in the inter tropical region is controlled by soil moisture .~ot by sc~,ii temperature because the soil temperatm'e does not fluctuate from one season to another by very many degrees. In the: higher latitudes, the same mean annual temperature .rr, eans .*.hat the soil is much warmer than the average in summer and ~uch colder in the winter and the growing season ma.v be co~rol led by both temperature and moisture. Therefore for inte,.'pre~.ations at the higher categoric levels that one uses on small scale maps it is necessa.ry ,'o make a distinction between ~he soils whose tempe~'~ture vary widely betweeL~ suborner and winter and soils which have the same temperature in summer or in winter.

The limi,~s of 5 degrees Celsius difference between summer and winte~ were proposed on the basis of an examination c,f the air temperatures at the two tropics. No criticisms were received before ,~,oil Taxonomy ,~,as pri~ted. However, it seems that probably the hyperthermic temperatures should have been inc!u~ed with the isohvperthermic temperatures for the basis of interpretations. This is a p~oblem that needs examination perhaps more generally and yet the tropic great groups which are defined by the difference between winter and summer temperatures probably should hav~ ,~ncluded the soils that have hyperthermic temperatures. The distinctions between soil temperature ,:lasses are shown at the family level but there are many sma~.l scale m~ps that cannot ~.,.se sol! families in the legend and if the temperatures are not in6'nated generally by the ~,ame of lhe n~ap unit in the taxonomy, then the temperature ~as to be ih;;'oduced as a phase. In general climatic ph;~e~ ar,~ i:i~practical because there is no universally acceptable clz.~L~ication of climate. In addition climatic maps are normally on v ~ small scale and cannot be t~ef~l for large scale maps and the relation between the air climate and the soil climat,: is qu;~te imperfect. There are not enough meteorologinal stations in the world ~.o sho,w the raL'~ shadow~ that exist in the mountainous areas.

From the point oi ~ view of soil genesis the soils whose growing seasons are controlled by temperature, have, in the fall of the year a cessation of plant - " grOWth and one has a f!usL~ of new foliage on the surface of the soil. The leaves of trees, the dry grasses, and so on. the crop residues, all provide !arg~ amou~ts of fresh organic matter at the ~oil surface. In the hmnid parts of ",_he inter tropical regior~s where there is no dry season and no control on the growing season by moisture or temperature, there are no large flushes of fresh organic matter. One find~ instead that the leaves drop a~. any month of the year in small numbers and there is a continuous accretion of organic litter at the surface but no large flush of new organic matter at the surface. In the inter tropica! regions where the growing season is co~trt~lled by moisture, the pla~ats stop growing when the rains stop and the leaves fall, the grasses die, and there is little difference in the flush of f r ~ h organic litter between the t copies and inter tropical regions. Therefore the tropic g~'eat groups are all defined ~ having a udic moksture regime rather than an ustic moisture regime. There seems to be ,~ difference in the genetic affects of a large amount of organic matter coraing over a short period and the same amount of organic matter coming over a full year. One sees differences between the soils of the humid tropics and the humid temperate regions that c~.n hardly be explained other than on -'.he basis of the kev leading effects of large amoua~.s of ~oluble organic materials coming within a short time and the same m o u n t s coming very e',lea.ly spaced over the year.

Question 43

You have often said ~ me ~nd I have often heard you state in meetings, that tax~ are not real things but are concepts. I think a lot of people h~ve continuing difficulty with that

k

- 3 3 -

Leamy Interview

s tatement because it req~,fires a IG~ of thought and it requires an understanding o f h o w taxonomy was developed. For the benefl~ for the rest of us here that don't have quite such a clear grasp of that statement, would you like to elabora:e for us?

Guy Smith:

Fcrhaps the root of the confusion aboui whether or not a taxon is a cGncept or a real thing lies in the custom we have of using the same name for a taxon such as Miami loam and using that name also for a unit on a map legend where we portray or try to portray the aerial extent of the ~oils that conform to our co'acepL~, of Miami loam. We are using ghe same name with two different meanings. The third one is that if we examine a pit and find a pedon that conforms to our concept of Miarrd loam, we say t~is ig Miami loam. It is not Miami loam. Miami loam is a concept not a real thing but we call it Miami loam because it has ~I1 the properties that correspond with our concept of Mia~d loaw~, f f Miami loam as a taxon or a real thing it would be impossible to change o::r concept of it },~ :.~,,,use it would be fixed by nature. Therefore, t~e original concept of Miami loam, whi,2h w ~ '.i'~.':~t of a soil developed in glacial .rnaleria~s, glacial t i l l glacial outwash, we could not have su;~d~vided Miami loam into something like the hundred or so soil series that were once caiied/v']iami loam.

There is no particular difference between the taxonomies of soiis and the living organisms. One can take a. pariicuiar plant and dry it and put it in an herbar,;um as a ,'-ype sample of particular species. However, the name of a species is binomial, it requires the name of the genus and of the pa:'ticular individuals that constitute a class witifin the genus. However, one may not put a genus in an herbarium. That is impossible. The tenets is obviously a concept or the botanist would not occa~,.Mnally r)~vise their classification and establish new genera. Many pedologists have had one introd',)cto~'y co,.trse in botany in which they had four or five lectures concerning taxonomy of l:)ban~. Everytb_iag, therefore, in plant taxonomy seems si~...ple to the~ but the problems in plant taxonomy is probably more difficult than those in soil taxon()my if you star: to delve into the ~qterature.

Question 44

In plant and animal taxo~iomi~ there has always been a rule of priority of name. In soil c!as~ific~- tions in the past, this r~le of 9riorit7 h,qs not been practical because the definitions of taxa have not been sufficicnti'~ precise. Now that we do have a Soil Taxonomy with precise definitiora;, should there be a rule of priority in terms of the names of taxa?

Guy Smith:

In the earlier classifications, there hv~ been confusion because the same names were used fo~" quiie different kinds of ~oiB in different countries. For example, the podzols of Russia were not the podzols of weste~'n Eurof.)e or Notch America. The difficulties arose from the lack of definitions that could be interprezed in the ~ame manner in differen, countries.

. ~ ¢ ~ It would be my I~rSoaai ore~ecen,,e that if o~e does not li~zc~ the particular taxa that are defined in Soil Taxonomy s~ch as a "fyp/c Hapluda!f and one wants to define another sort of tax,.~n that would include some Typic Hapludalfs and ~ rhaps some MollisoL~ and so on , one should take a new name for --'~ ~ s u ~ a taxon rather than retaining the presently defined name and presenting a new definition for ~ e s~m:, name. Frora my ow~ experience, I would ~iuch prefer to present a new name for a ~aew ~axon. Any decL~ion to ~i~e priority to nomenclature ha pedology would require some sort of intonational agr~:ment by the Int~.a.tional Soil So;once ~-ciety. I am only one m ~ m ~ r cf ti'~t So, iety and I may no t : .... d.,.t,,t6 what i~ decides to do but to this date ~ e r e have never b-~en to my knowledge proposals to give priority to particular

- : - x

: ~ \ 7 " .¸ ~ . •

k - - - ~. ' 2

- 34 -

I I

Leamy lnterviev,

names. The FAO/UNESCO .~,,o.,,.~,..._,.~ for the soil map of the world uses a few of Soil Taxonomy's names but for the most part ~:,ses its own names. While the legend for a soil map such as the FAO/UNESCO soil ma~ of the world is not itself z taxt~nomy still the units are named as though it were a taxonomy. If one wants to give priority to nomenclature then the FAO,/UNESCO legend woul~ h~v,~, been impelled to use the soil names in Soil Taxonomy but for the legend this might not have neees,~arily been convenient or useful. It would be my own opinion that i f one establishes a system of priority it would be a serious m~sta~ce at this moment because it would tend to freeze.,, the ,~xisfing names and would proven ~, changes in defL-,itions.

Question 45

The next question is a common one among pedologists working on floodplains with En|isols and Incep¢isols ~nd concer~ the difficulty of distinguishing a cambic horizon where the soil has an aquic moisture regime. The question is directed both from Thailand and from New Zealand and asks if ~h~.re Ls any better method of distinguishing the degree of alteration for soils having a aquic moisture rcz~i~'ae.

Guy Sr~ith:

Whi!e in New Zealand, you macie ~ proposal based on New Zealand experience for an 8ddition,'~l criterion for identifying szlch a cambic horizon. The proposed change reads: In Soil T~xonom.v, page 34, ?eft column, starting ot; the 9th line from the bottom where it reads. ~-0-~ evidence of alteration" dele'¢ the balance of the paragraph and substitute the following column. "One evidence of pedogenic at.teraiio~ is the formation of iron ol iron manganese concretions of swad or coarser size. Although ~oltlcs of low chromo may develop before the flood waters thai deposit the alluviz~m have receded, concretions of iron or ir,;n mangmzese forms slowly. But, mottles can grade into concretions, and son~:e limi: mgst be set between them. A concretion, to be diagnostic o1' alteration, must be cemented to the point lhat it does no~ break down under normal particle size dispersion mei,hods, and can't be separated by screening. One concretion in ./0 grams of soil would not seem significan:f in a poorly drained soil, bui i f something like I /4 percent of the soil, by weight, cons~"st.~ of concretions, the evidence of alteration adequate for a cambic horizon is satisfied. Concretiorz may be, distinguished from dark colored minerai grains by their fragility under a gentle pressure, the presence of MoO 2, or magnetic properties that change after ignition."

Change item 4.a.l., pa,'fe 36. left ¢olur~'~n. to read: "Have at least I /4 percent by weight of iron or iron manganese concretion~ of sand size or larger wilhin the surface 50 cm that do not disperse oa overnight shaking m h~:,xametc, phosphate.

References cited

Cain . ClJ ...............................

C l ~ e , M. G . 1949. Basic p,,'~nci¢!~,.s e~e soi.~ classification. Soil ~.'i. 7 :81-91.

- 35 -

Leamy Interview

Cline, M. G. 1961. The changing mode! of soil. Soil Sci. Soe. Am.. Proe. 25:442-446.

Cl/ne, M. B. 1977. Historic,q~. highfights in soil genesiz, morphology and classification. Soil Sci. Soc. Am. ~. 41 250-254.

Mill, John Stuart. 1891. A syste, m of logic. 8th Edition. Harper and Bros., New York 659 pp.

Smith, G. D. 1965. Lectures on soil classification, Pedologie, Special Bull. Ghet~t, Belgium,

Soil Survey Staff. 1975. Soil taxonomy, a basic system of soil classification for makivg and interpreting soil ~;rveys. USDA. Agric. "~ • rzandb.c~.k No. 4~6. 754 pp. U.S. Gov't. Printing Office, Waslhing~en, D. C.

- 36 -